9th International Conference on Sandwich Structures

ICSS 9

G. Ravichandran (Editor)

IMPACT PROPERTIES OF CORRUGATED COMPOSITE SANDWICH CORES

S. Kazemahvazi*, J. Kiele*, B. Russel†, V. Deshpande† and D. Zenkert* * KTH, Department of Aeronautical and Vehicle Engineering

SE-100 44, Stockholm, Sweden e-mail: sohrabk@kth.se, web page: http://www.ave.kth.se

† Engineering Department, University of Cambridge

Trumpington Street, Cambridge CB1 1PZ, UK

Key words: Corrugated core, micro-inertia, impact, sandwich, high performance core.

Abstract. The out-of-plane impact properties of corrugated carbon fibre composite cores have been investigated experimentally. Cores with slender core members show dynamic strengthening of approximately 12 times compared to quasi-static experiments whereas cores with stocky core members show a dynamic strengthening of approximately 2 times. Carbon fibre corrugated cores have superior compressive performance in the density range of 100 kgm-3 to 300 kgm-3. 1 INTRODUCTION

Sandwich structures, made of thin high performance face sheets separated by a thick low-density core, have superior bending stiffness and strength compared to a monolithic structure of equal weight. Traditionally, the low density cores of the sandwich structure are made of polymeric foams of various relative densities. Foam materials are however bending dominated structures, and the material is thus not fully utilized, which results in a lower weight specific performance compared to a stretching dominated structure [1]. During the past decade there has been an effort to develop novel core topologies that are periodic and stretching dominated for one or several different loading scenarios [2, 3]. The core topologies can be divided into two main categories, prismatic cores and lattice truss cores. Examples of prismatic cores are square honeycombs [4, 5], corrugated cores and diamond configuration cores [6-8]. Lattice truss cores typically consist of pyramidal, tetrahedral, kagomé or textile configurations [9-13]. Since these periodic core configurations are stretching dominated the predominant mode of failure is onset of elastic or plastic buckling, especially for core configurations with a low relative density. In order to increase the resistance to buckling without adding substantial weight, variants of tubular lattice truss cores [14-16] and hierarchical core concepts [7,17,18] have been developed and show significant increase in strength compared to its monolithic counterpart. Figure 1 summarises the experimental compressive strengths for a majority of sandwich core materials available in the literature today. In addition to good quasi-static performance, sandwich structures that are used in military vehicles or naval ships need to have good resistance to blast and ballistic loading conditions. Several studies have investigated the behaviour of prismatic and

Sohrab Kazemahvazi, Joern Kiele, Benjamin Russel, Vikram Deshpande and Dan Zenkert

2

lattice truss cores when subjected to ballistics [19, 20] and blast loading [21-27]. Most studies report an increase in blast performance for sandwich configurations compared to monolithic structures, with square honeycombs having the best out-of-plane performance, corrugated and diamond cores having high longitudinal stretching performance and lattice trusses having competitive performance at low core densities.

Figure 1: Experimental data for the out-of-plane compressive strength of the majority of available sandwich core configurations up to date.

Dynamic loading scenarios of the aforementioned periodic cellular cores differ from the quasi-static loading case in three fundamental ways. First, the constituent material of the structure may show strain rate dependence. Second, since the cellular cores are buckling dominated, inertial effects can delay the onset of buckling and/or change the wave length of the buckling mode. Finally, propagation of elastic, plastic and bending waves can be transmitted through the core which can affect the macroscopic properties of the structure. Ferri et al [28] conducted a comprehensive study on the dynamic behaviour of stainless steel I-core configurations and found strong inertial influence resulting in a substantial decrease of the buckling wave length as the loading rate increased. Tilbrook [24] et al investigated the dynamic crushing of stainless-steel corrugated and Y-core configurations. Both core topologies showed strong inertial stabilisation as the loading rate increased which resulted in substantial increase of the collapse strength and a decrease of the buckling wave length. At impact velocities below 30 m/s, the stresses measured at the front and rear faces of the sandwich were approximately the same. This indicates a state of axial equilibrium as the core collapses. As the impact speed increased, wave propagation effects played a dominant role and the measured peak stresses at the front face exceeded that of the rear face. In fact, the peak stress of the rear face remained approximately constant (and equal to the low speed impact peak stress) while the peak stress of the front face continued to increase as the loading rate increased (up to a certain limit).

Although the dynamic response of metallic lattice truss and prismatic cores have been

101

102

103

104

105

10−1

100

101

102

Foam Rohacell Core IG and WF Evonik RohacellFoam Dyvinicell HP Grade DIAB GroupWood Cores [Reid1997, DIAB ProBalsa]Composite Monolithic Corrugated Core [Kazemahvazi2009]Composite Hierarchical Corrugated Core [Kazemahvazi2009]Composite Pyramidal Truss [Finnegan2007]Composite Square Honeycomb [Russel2008]Composite Square Hierarchical Honeycomb [Côté2009]Composite Folded Core [Heimbs2008, Heimbs2009a]Steel Pyramidal Truss [Lee2006, Rathbun2006, Zok2004]Steel / Copper Tetrahedral Truss [Lim2006, Wang2003]Steel / Copper Kagomé Truss [Lim2006, Lee2006, Lee2007, Wang2003]Steel Prismatic Lattice [Tilbrook2007]Steel Corrugated and Diamond Lattices [Côté2006]Steel Woven Core [Zupan2004]Aluminium Honeycomb [Plascore PAMG−XR15056]Foam Aluminium Alloy [Koza2004/McCullough1999]Aluminium Egg Boxes [Zupan2003]Aluminium 3−D Truss [Wallach2001]Aramid Fiber Honeycomb [Plascore PN1]Aramid Paper Folded Core [Heimbs2008, Heimbs2009a, Kintscher2007]

Eqvivalent core density (kgm-3)

Out

-of-

plan

e co

mpr

essi

ve st

reng

th (M

Pa)

Sohrab Kazemahvazi, Joern Kiele, Benjamin Russel, Vikram Deshpande and Dan Zenkert

3

thoroughly explored there has been little research done on the fibre composite counterpart. In this work the out-of-plane compressive dynamic behaviour of a prismatic composite core is explored with a thought future application as high performance and multifunctional cores in naval ship hulls. Experiments are performed at quasi-static loading rate, low speed impacts and high speed impacts. The low- and intermediate speed impacts would resemble a loading scenario of a ship collision and hull slamming loads while the high speed impacts simulate blast loading scenarios. Especially the effects of inertial stabilization are studied and its effect on the out-of-plane compressive strength of the core.

2 MATERIALS AND EXPERIMENTS PROGRAMME The corrugated cores were made out of low temperature curing epoxy prepreg system

reinforced with carbon fibres [29]. Both unidirectional and weave configuration was used. Each material configuration was tested in tensile and compression according to ASTM D3039 and ASTM D3410 respectively, and the strengths and modulus were measured accordingly. A summary of the material properties is found in table 1. The core members were bonded to the face sheets using Araldite 420A/B epoxy adhesive.

Type Ply thickness Fibre architecture

TR50 ~0.26 mm UD 142 GPa 650 MPa 2030 MPa

RC203P ~0.18 mm 0/90 plain weave 75 GPa 500 MPa 1380 MPa

Table 1: Experimentally measured material properties of SE-84LV low temperature curing

prepreg system. Index 1 refers to the properties in the fibre direction, 1ˆ cσ the compressive

strength and 1ˆ tσ the tensile strength.

A flat sheet of monolithic composite laminate or a sandwich laminate was manufactured by stacking the desired amount of plies and curing the prepreg at 120ºC for 1h. The sheet was cropped into several smaller sheets of the size of a core member and then mounted into a gluing fixture, see figure 2a. Side supports were bonded to the bottom face sheet to give lateral support to the core members. In a plate or a beam structure this would not be necessary since the core is periodic and each core member would then be supported by its neighbouring core member. Subsequent to the bonding of the bottom face sheet, the top face sheet was bonded to the core using a fixture that ensures that the face sheets are parallel, see figure 2b.

Figure 2: (a) Gluing fixture for bonding the bottom face sheet to the core members and (b) a fixture to bond the top

face sheet to the core members.

A schematic view of the unit cell configuration is given in figure 3 and a summary of the core configurations that are analyzed within this work is summarised in table 2.

(a) (b)Monolithic

1cσ̂ 1tσ̂1E

Sohrab Kazemahvazi, Joern Kiele, Benjamin Russel, Vikram Deshpande and Dan Zenkert

4

Mono1 Mono2 Mono3 Mono4 Core density (kg/m3) 35 70 55 210

ω (°) 70 70 70 70 t (mm) 0.26 0.52 0.46 1.81

Material TR50 TR50 RC203P RC203P l1 (mm) 35 s (mm) 15

Table 2: A summary of the different manufactured core configurations.

Figure 3: Unit cell of a corrugated core. The left core member shows a monolithic version and the right core member

shows a sandwich version.

For the monolithic configuration four different densities were tested, two with unidirectional core members and two with plain weave fibre architecture as core member.

Quasi-static axial compression of a single unit cell was performed in a standard screw-driven Instron machine with a 100 kN or a 30 kN load cell depending on the expected failure load of the unit cell. The axial compression displacement of the unit cell was measured using an extensometer that was mounted between the rigid steel plates that compress the unit cell.

The dynamic compression experiments were performed by firing rigid projectiles using a gas gun set-up and the load response was measured using a Kolsky bar [30]. Experiments were performed at projectile impact velocities, Vp, ranging from 5 ms-1 to 50 ms-1. The mass of the projectile, mp, was chosen so that a constant velocity crush of the unit cell was obtained. High speed photography confirmed that the velocity of the striker remained approximately constant. Readers are referred to Radford et al [31] for further details including the calibration procedure.

3 SUMMARY OF EXPERIMENTAL FINDINGS

3.1 Quasi-static experiments The compressive failure modes of the different core configurations are presented in figure 4.

s z

x

ω

l1

tc t side support

mp, Vp

Sohrab Kazemahvazi, Joern Kiele, Benjamin Russel, Vikram Deshpande and Dan Zenkert

5

Mono 1-3 are all slender structures and the failure is governed by elastic buckling of the core members. Mono 4 is sufficiently stocky to prevent elastic buckling and the failure is governed by the compressive material strength of the core member. The mode of failure for mono 4 is a combination of delamination and fibre micro buckling as is seen in figure 3d. In the proceeding sections mono 1-3 will be referred to as buckling dominated structures and mono 4 will be referred to as compression failure dominated structure.

Figure 4: Quasi static failure modes of (a) Mono 1, (b) Mono 2, (c) Mono 3 and (d) Mono 4.

3.2 Dynamic experiments A summary of the normalised dynamic strength of the different core configurations, impacted

at different velocities, is given in figure 5. The buckling dominated structures, mono 1-3, show significantly larger increase in strength as function of impact velocity compared to the compression failure configuration. Further, the increase in dynamic strength is the highest for the configurations with the lowest bending stiffness, mono 1 and mono 3. In the proceeding section these findings will be discussed and compared to the observations made using high speed photography.

3.2.1 Buckling dominated configurations

The buckling dominated structures show a significant increase in strength with increasing impact velocities. Typically there is a sharp increase in strength which is followed by a plateau of approximately constant strength as the impact velocity increases. The knee of the plateau is at approximately 10 ms-1 for the low bending stiffness configurations, mono 1 and 3, and at 20 ms-1 for the mono 2 configuration.

Figure 6 shows photographs of the mono 1 structure at two nominal displacements for

different impact velocities. At low speed impacts, 5 ms-1, the core members deform into a buckling mode shape similar to the first and second buckling mode of a clamped-clamped strut. This buckling mode shape is immediately followed by splitting failure of the core member. At 10ms-1 the buckling wave length of the core members continues to decrease and consequently the failure load increases. The decrease in wave length and increase in failure load is due to the inertia effects that arise when the core members are forced into a certain buckling mode at high speeds.

(a) (b)

(c) (d)

Sohrab Kazemahvazi, Joern Kiele, Benjamin Russel, Vikram Deshpande and Dan Zenkert

6

Figure 5: Normalised dynamic strength as function of impact velocity for the four different core configurations.

At impact velocities that are on the plateau, 27 ms-1 and 50 ms-1, the core members are fully stabilised due to inertia and the failure is set by the compressive strength of the material. This strength is about 12 times the quasi-static strength of the structure. The shift in failure mode is clearly elucidated in figure 6 where the initial failure, of both the 27 ms-1 and 50 ms-1, occurs while the core members are considerably straight.

Figure 6: High speed photography of impacts at various speeds on the mono 1 structure.

To ensure that there were no three dimensional buckle modes of the core members,

perspective view photographs were taken. Figure 7 shows a series of perspective view photographs of the mono 1 structure impacted at 27 ms-1. Immediately after impact the core members deform in a very short wave length mode and areas of splitting failure can be observed. The deformation mode of the core member is the same over the width of the unit cell and it can be seen that a band of bending failure forms close to the bottom boundary.

Nominal strain:

0.015

0.030

5 m/s 10 m/s 27 m/s 50 m/s

Mono 2 M̂ono 3 M̂ono 4 ˆ

Impact velocity, Vp (ms-1)

D

Q-S

σ̂σ̂

Sohrab Kazemahvazi, Joern Kiele, Benjamin Russel, Vikram Deshpande and Dan Zenkert

7

Figure 7: Perspective view of the development of failure of the mono 1 configuration impacted at 27 ms-1.

The failure behaviour of the mono 3 structure is shown in figure 8. The mono 3 structure behave similar to the mono 1 structure with a few exceptions. The decrease in the buckle wave length is observed for the mono 3 structure as well but the proceeding failure mechanism is however not splitting but fibre failure which can be observed in figure 8. This is since the weave architecture of the laminate prevents it from splitting and hence fibre failure occurs once the curvature of the core member becomes sufficiently large. At 10 ms-1 the buckling wave length of the structure is very small and as the impact velocity increase the structure is fully stabilised and compressive material failure is observed.

Figure 8: High speed photography of impacts at various speeds on the mono 3 structure.

The perspective view of the mono 3 configuration is shown in figure 9 and the photographs confirm the 2D deformation state of the core members.

Figure 9: Perspective view of the development of failure of the mono 3 configuration impacted at 27 ms-1.

Nominal strain:

0.015

0.030

5 m/s 10 m/s 27 m/s 50 m/s

Sohrab Kazemahvazi, Joern Kiele, Benjamin Russel, Vikram Deshpande and Dan Zenkert

8

3.2.2 Compression failure dominated structures

The compression failure dominated configuration show considerably smaller increase in strength as the impact velocity increases. Since the structure fails in compression failure at quasi-static loading rate, only small effects of inertia stabilisation are present in the dynamic loading scenario. Hence the observed dynamic strengthening is mainly due to the material strain rate sensitivity. At an impact speed of 50 ms-1, the strain rate over a core member is approximately 1000 s-1 and the dynamic strength is approximately twice as high as the quasi-static strength which is in agreement with studies on the compressive material properties of carbon fibre composites [32, 33].

Figure 10 shows high speed photographs for three different impact speeds. At low speed, delamination and fibre failure occurs over the entire core member and as the speed increases the failure gets more localised and occurs closer to the front face (impacted face) of the specimen.

Figure 10: High speed photography of impacts at various speeds on the mono 4 structure.

4 CONCLUDING REMARKS The dynamic response of four different corrugated composite cores has been investigated

experimentally by impacting unit cells of the core with projectiles at impact velocities ranging from 10ms-1 to 50 ms-1. The main observations are summarised as follows.

• Low relative density cores show significant dynamic strengthening (up to 12 times

the quasi-static strength). The strengthening is primarily due to inertial stabilisation of the core members.

• High relative density cores are compression failure dominated at quasi-static loading and hence they show small inertial stabilisation when loaded dynamically. The dynamic strengthening, which is in the order of 2 at strain rates of 1000s-1, is thus primarily due to material strain rate strengthening.

• Monolithic corrugated carbon fibre composite cores, together with carbon fibre composite square honeycomb cores, show highest compressive strength at core densities between 100kgm-3 to 250 kgm-3.

Nominal strain:

0.020

0.040

10 m/s 27 m/s 50 m/s

Sohrab Kazemahvazi, Joern Kiele, Benjamin Russel, Vikram Deshpande and Dan Zenkert

9

ACKNOWLEDGMENTS The financial support for this investigation has been provided by The Office of Naval

Research (ONR) through programme officer Dr. Yapa D.S. Rajapakse (Grant No. N00014-07-1-0344).

REFERENCES [1] Deshpande VS, Ashby MF, Fleck NA. Foam topology bending versus stretching

dominated architectures. Acta Materialia, 2001;49,1035-1040. [2] Wadley HNG. Multifunctional periodic cellular metals. Philosophical Transactions of

the Royal Society A. 2006;364,31-68. [3] Wadley HNG. Fleck NA, Evans AG. Fabrication and structural performance of

periodic cellular metal sandwich structures. Composite Science and Technology. 2003;63,2331-2343.

[4] Coté F, Deshpande VS, Fleck NA, Evans AG. The out-of-plane behaviour of metallic honeycombs. Materials Science & Engineering 2004;380,272-280

[5] Russel BP, Deshpande VS, Wadley HNG. Quasistatic deformation and failure modes of composite square honeycombs. Journal of mechanics of materials and structures 2008;3,7.

[6] Kazemahvazi S and Zenkert D. Corrugated all-composite sandwich structures. Part 1: Modeling. Composite Science Technology, 2009;69;913-919

[7] Kazemahvazi S, Tanner D and Zenkert D. Corrugated all-composite sandwich structures. Part 2: Failure mechanisms and experimental programme. Composite Science Technology, 2009;69;920-925

[8] Coté F, Deshpande VS, Fleck NA, AG Evans. The compressive and shear responses of corrugated and diamond lattice materials. Int J Solid Struct 2006;43,6220-6242.

[9] Kooistra GW, Deshpande VS and Wadley HNG. Compressive behaviour of age hardenable tetrahedral lattice truss structures made from aluminium. Acta Materialia 2004;52,4229-4237.

[10] Kooistra GW and Wadley HNG. Lattice truss structures from expanded metal sheet. Materials and Design 2007;28,507-514.

[11] Biagi R and Bart-Smith H. Imperfection sensitivity of pyramidal core sandwich structures. International Journal of Solids and Structures 2007;44;4690-4706

[12] Hyun S, Karlsson AM, Torquato S and Evans AG. Simulated properties of Kagomé and tetragonal truss core panels. International Journal of Solids and Structures 2003;40;6989-6998

[13] Finnegan K, Kooistra GW, Wadley HNG and Deshpande VS. The compressive response of carbon fiber composite pyramidal truss sandwich cores. Int. J. Mat. Res. 2007;98,12.

[14] Queheillalt DT and Wadley HNG. Pyramidal lattice truss structures with hollow trusses. Material Science & Engieneering A 2005;397,132-137.

[15] Rathbun HJ, Zok FW, Waltner SA, Mercer C, Evans AG, Queheillalt DT and Wadley HNG. Structural performance of metallic sandwich beams with hollow truss cores. Acta Materialia 2006;54,5509-5518

[16] Lee BK and Kang KJ. Compressive strength of tube-woven Kagome truss cores. Scripta Materialia 2008;60,391-394.

[17] Kooistra GW, Deshpande VS and Wadley HNG. Hierarchical Corrugated Core

Sohrab Kazemahvazi, Joern Kiele, Benjamin Russel, Vikram Deshpande and Dan Zenkert

10

Sandwich Panel Concepts. Journal of Applied Mechanics 2007;74,259-268. [18] Coté F, Russell BP, Deshpande VS and Fleck NA. The through thickness

compressive strength of a composite sandwich panel with a hierarchical square honeycomb sandwich core. Journal of Applied Mechanics 2009;76,.

[19] Yungwirth CJ, Radford DD, Aronsson M, Wadley HNG. Experiment assessment of the ballistic response of composite pyramidal lattice truss structures. Composites Part B: Engineering. 2007

[20] Yungwirth CJ, Wadley HNG, O’Connor JH, Zakraysek AJ, Deshpande VS. Impact response of sandwich plates with a pyramidal lattice core. International Journal of Impact Engineering. 2007;35;920-936.

[21] Rathbun HJ, Radford DD, Xue Z, He MY, Yang J, Deshpande VS, Fleck NA, Hutchinson JW, Zok FW, Evans AG. Performance of metallic honeycomb-core sandwich beams under shock loading. International Journal of Solids and Structures. 2006;43;1746-1763.

[22] Rubino V, Deshpande VS, Fleck NA. The dynamic response of clamped rectangular Y-frame and corrugated core. European Journal of Mechanics. 2009;28;14-24.

[23] Rubino V, Deshpande VS, Fleck NA. The dynamic response of end-clamped sandwich beams with a Y-frame or corrugated core. International Journal of Impact Engineering. 2008;35;829-844.

[24] Tilbrook MT, Radford DD, Deshpande VS, Fleck NA. Dynamic crushing of sandwich panels with prismatic lattice cores. International Journal of Solids and Structures. 2007;44;6101-6123.

[25] Radford DD, McShane GJ, Deshpande VS, Fleck NA. Dynamic compressive response of Stainless-steel Square Honeycombs. Journal of Applied Mechanics. 2007;74;658-667.

[26] Dharmasena KP, Queheillalt DT, Wadley HNG, Dudt P, Chen Y, Knight D, Evans AG and Deshpande VS. Dynamic compression of metallic sandwich structures during planar impulsive loading in water. European Journal of Mechanics. 2010;29;56-67.

[27] Fleck NA and Deshpande VS. The resistance of clamped sandwich beams to shock loading. Journal of Applied Mechanics, 2004;71,1-16.

[28] Ferri E, Antinucci E, He MY, Hutchinson JW, Zok FW, Evans AG. Dynamic buckling of impulsively loaded prismatic cores. Journal of Mechanics of Materials and Structures, 2006;1,8.

[29] Gurit Ltd, Datasheet SE-84LV, Newport, Isle of Wight, United Kingdom, http://www.gurit.com

[30] Dharan CKH, Hauser FE, Testing techniques based on the split Hopkinson bar. Experimental Mechanics, 1970;10, 70-376.

[31] Radford DD, Deshpande VS, Fleck NA. The use of metal foam projectiles to simulate shock loading on a structure. International Journal of Impact Engineering. 2005;31,9,1152-1171.

[32] Hsiao HM, Daniel IM. Strain rate behavior of composite materials. Composite Part B. 1998;29B;521-533.

[33] Guedes RM, de Moura MFSF, Ferreira FJ. Failure analysis of quasi-isotropic CFRP laminates under high strain rate compression loading. Composite Structures. 2008;84;362-368.