Document [original]

Original article

Experimentally examining the

mechanical behaviour of nap-core

sandwich material – A novel type

of structural composite

Giap X Ha

1,2

, Andreas Bernaschek

1,2

and Manfred W Zehn

1,2

Abstract

Experimental investigation of the nap-core sandwich is presented in detail, in which the nap-core is based on knitted

fabric, being impregnated with a thermoset resin, formed to create cup-shaped naps, and stabilized to assume a per-

manent 3D contour. The material is a novel type of lightweight sandwich-structured composite which has good specific

strengths and possesses various properties crucial for engineering applications, but its exploitation is still restricted due

to insufficient research and understanding. The sample preparation is first described, being followed by the test imple-

mentation and outcomes. The results obtained from typical tests demonstrate high performance of the nap-core

sandwich samples under different cases of loading. They also reveal the sandwich’s essential behaviours which are

similar to those of a common shell structure, giving it a great potential of being computationally modelled with finite

element software.

Keywords

Sandwich structures, thermosetting resin, polymer fibres, structural composites, textile reinforcements, mechanical

behaviour

Introduction

The nap-core sandwich composite

For decades, lightweight sandwich-structured compo-

sites (also called core materials) have been in an

increasing demand in engineering applications on

account of their high strengths and stiffness to the den-

sity, especially the compressive strength and the flexur-

al strength.

1–4

Basically, the material is created by

sandwiching a core between two stiff skins using

some kind of bonding. In the industry of manufactur-

ing aerospace and aircraft, non-metal sandwich mate-

rial is used to fabricate most of lining elements in which

the outer layers usually are impregnated phenolic resin

– glass fibre laminates and the core is honeycomb or

foam. When parameters of the core change, the behav-

iour of the sandwich changes as well. That results in a

wide range of properties for numerous applications.

5–7

However, sandwiches with honeycomb or foam core

have drawbacks which are limited drapability and

closed inner structure. The former causes a difficult

employment of the sandwich to curved surfaces, and

the latter probably make accumulation of condensation

water that increases weight and reduces the mechanical

properties of the sandwich.

8–10

In that sense, the nap-

core is derived from textiles to overcome the recent

problems of the other cores. A typical nap-core has a

3D shape obtained from deep drawing, curing, and

cooling a resin-impregnated 2D knitted fabric (see

Figure 1).

The most successful fibre materials are

thermoplastic polymers, aramid, glass, cellulose,

basalt, and hybrid fibre. These fibres are non-toxic

and strongly resistant against heat, solvents, hydrolysis,

Field of Computational and Structural Mechanics, Institute of Mechanics,

The Technical University of Berlin, Berlin, Germany

InnoMat GmbH, Teltow, Germany

Corresponding author:

Giap X Ha, Institute of Mechanics, Strasse des 17. Juni 135, Berlin 10623,

Germany.

Email: [email protected]lin.de

Journal of Reinforced Plastics and

Composites

2019, Vol. 38(8) 369–378

!The Author(s) 2018

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/0731684418820437

journals.sagepub.com/home/jrp

and oxidizing agents. Textiles made of them also prove

to be highly durable and tough. Nevertheless, many of

those materials have low elongation of break (around

4%) while the production of the nap-core necessitates

the textile to stretch up to 250%. Therefore, fabrics fab-

ricated by knitting technique are used as they are incred-

ibly suitable for creating deep-drawn shapes without

local fractures or crimps.

The best materials for the matrix are thermoset resins

Cyanate Ester, Epoxy and Phenol Formaldehyde for

their working stability and health safety (good fire-

smoke-toxicity standard). Beyond that, they are reason-

able at price for commercial production and own a fast

cross-linkage and low curing time. In practice, the select-

ed resins have already shown good basic properties plus

high compatibility with plasticizer and retardants as well

as their excellent ability of wetting fibrous material.

Geometrically, a nap-core is a combination of numer-

ous periodic naps. To add bending stiffness, two hard

skins are attached to the top face and bottom face of the

nap-core to form a sandwich as shown in Figure 2.

Although the nap-core sandwich has specific strengths

not as high as those of the honeycomb at the same den-

sity, it offers many other desirable properties. In partic-

ular, the nap-core sandwich is rather flexible, so it can be

bent over complex surfaces. The hollow fabric structure

of the core allows good drainage and ventilation to fluid,

and the slots between naps facilitate the integration of

cables or tubes into the structure.

In addition, altera-

tions of constituent materials and underlying geometries

of the nap-core help to generate various types of the

sandwich with adjustable properties to fit different

requirements. Depending on the layout of naps, nap-

cores are divided into two main types, single-sided nap-

core and double-sided (symmetrical) nap-core. Of them,

the former is more convenient for handcraft, and the

latter is more suitable for automatic production.

In comparison with common engineering textiles,

the nap-core sandwich also possesses many advantages

making it special. Firstly, the use of thermoset resin

helps to repair small fractures and reduce wrinkles

and disorders on the knitted fabric. Secondly, the form-

ing and stabilizing steps keep an enduring shape of the

core, giving it a necessary height without taking up

more material. Thirdly, after cooling, the matrix will

lock all fibre positions and prevent transposition of

yarns so that eliminate the problem of inter-tow slid-

ing.

Overall, the nap-core sandwich possesses high

strength to density (specific strength), especially under

compression and bending.

Although containing a large number of merits, the

nap-core sandwich composite is currently not as reli-

able as honeycomb because it is still little understood.

Its applications are almost interior decoration and

cover in means of transportation whilst it is highly

potential to be used as load bearing components.

Thus, research of the material needs more attention

as only that will bring it a deserved development. In

this article, the authors will present an experimental

investigation of the nap-core sandwich’s mechanical

behaviours. The other investigation – modelling based

on finite element method – is likely to confront several

difficulties due to the complicated underlying structures

of the nap-core, and it will be present in a sepa-

rate article.

Recent studies

Although being patented in 1977, the nap-core sand-

wich composite is relatively innovative with only a few

years of mass production and development, so insight

into it is limited. There have been a number of reports

made by Monika et al. in 2006,

Bernaschek et al. in

2011,

and Gerber in 2017

describing its fabrication

Figure 1. Photos of a typical nap-core (left) and one of its naps (right).

Figure 2. Scheme of a representative nap-core sandwich.

370 Journal of Reinforced Plastics and Composites 38(8)

procedure, categories and properties in detail.

However, there is only one article published by

Gerber et al. in 2016 informing the result attained on

inspecting dynamic behaviour of a symmetric nap-core

sandwich under impact load.

This shortage explains

inadequate employment of the nap-core sandwich in

engineering applications despite its advantageous prop-

erties in all physical, chemical, and mechanical aspects.

On purpose of obtaining a more comprehensive

understanding on the nap-core sandwich composite,

the author has conducted a typical mechanical experi-

ments on samples of two types of the material. The

tests are compression, shear, three-point bending, and

four-point bending. The samples, implementation, and

results will be given in the next sections.

Comparison between nap-core sandwich and hon-

eycomb sandwich

To have a deeper understanding of the nap-core sand-

wich, a comparison between it and a honeycomb sand-

wich with similar boundary dimensions has been

carried out. For both kinds of sandwiches, the outer

sheets are the same kinds of laminate composed of

glass fabric and Phenol formaldehyde resin. The exam-

ined nap-core’s knitted fabric is made of Aramid

Hybrid for the yarns and Phenol formaldehyde for

the matrix, of which the naps are triangularly

arranged and have an average diameter of 6 mm. The

mass-to-volume ratio of the nap-core sandwich sample

is 55 kg/m

. The comparative honeycomb is made

of paper, having a cell width of 3.2 mm and a mass-

to-volume ratio of 48 kg/m

. The comparison is on

compressive behaviours. The samples are illustrated

in Figure 3, and the resulting behaviours of the two

sandwich samples are presented in Figure 4.

The experiments demonstrate that the honeycomb

sandwich has better strengths and moduli in general,

but the nap-core sandwich holds an outstanding advan-

tage. When the damage happens, the strength of hon-

eycomb sandwich descends abruptly and its force

plunges, whereas the force of the nap-core sandwich

goes down slowly with only a modest slope. This fea-

ture, in addition to great ventilation and an easy inte-

gration of wires, make the nap-core sandwich a good

selection in numerous applications.

14,15

Samples and experiments

Samples

Two types of nap-core and their sandwich composites are

going to be introduced and examined in this article, which

are P1-10 (single-sided nap-core) and P2-8 (double-sided

Figure 3. The honeycomb (left) and the nap-core (right) of the comparison. Source: Kunststoffe international GmbH.

Figure 4. Compression to a sandwich with nap-core (left) and honeycomb (right). Source: Kunststoffe international GmbH.

Ha et al. 371

nap-core). For the names of the nap-core types, “P”

stands for Phenolic resin; “1” denotes single-sided; “2”

denoted double-sided; “10” means 10 cm height; and

“8” means 8 cm height. They are among the most

common categories and different at all major elements:

constituent materials, knitting patterns, and geometries.

In other words, they are to some extent offer an initial

overview on the nap-core sandwich-structured composite.

Details are presented in Figure 5 and Table 1.

Sandwich samples are prepared following a stan-

dardized production procedure. At first, a 2D sheet

of knitted fabric is pre-impregnated with a thermoset-

ting phenolic resin to form a mixture in which the resin

embeds the fabric inside to create a wet mixture of

fabric and resin. The mixture is then laid between

two halves of a pin mould to give the knitted sheet a

3D shape (with a height of 5–10 mm) as a combination

of periodic-distributed identical cone-shaped naps.

Afterwards, the mould with the sample inside is cured

at 140C for 4–6 h and cooled down at room temper-

ature in a similar time. Henceforth, the nap-core takes

up a stable shape, and the yarns no longer slide to each

other. In the next step, the sandwich is completed by

bonding the stabilized nap-core with two outer face

sheets. Finally, the samples are cut from the big sand-

wich panel to desired dimensions.

As stated in the introduction, three usual experi-

ments (e.g., compression, shear, and four-point bend-

ing) are going to be implemented for the inquiry of the

sandwiches’ mechanical behaviours. The experimental

standards are D3410M – 03 for compression, DIN 53

294 for shear, and DIN 53 293 for four-point bending.

The sample dimensions (length width) are 5 5cm

for compression test, 20 5 cm for shear test,

and 40 5 cm for four-point bending test (refer to

Figure 6). The samples used for the current research

comprise of materials as displayed in Table 1. For

every kind of test, the number of the samples of each

nap-core sandwich type is always equal to 5.

Sample fixations

Samples are sized and installed as in the standards.

The schemes of fixtures are shown in Figures 7 and 8 in

Table 1. Specifications of the nap-core sandwich types used for the experiment.

Nap-core type Material

Boundary

height (mm)

Fabric

thickness (mm)

Volume weight

(kg/m

)

P1-10 55% fibre (90% Nomex þ10% Polyester) þ45% Phenolic resin 10 0.58 83

P2-8 50% fibre (80% Aramid þ20% Polyester) þ50% Phenolic resin 8 0.45 41

Figure 5. Nap-core types and P1-10 (left) and P2-8 (right).

Figure 6. Samples of the experiments: (a) four-point bending, (b) compression, (c) shear.

372 Journal of Reinforced Plastics and Composites 38(8)

which specific velocity of the load generator’s head for

each testing case is 10 mm/min for compression or four-

point bending, and 1 mm/min for shear or three-

point bending.

Results and discussions

Results

In this section, the results are shown and discussed. The

relationship between the applied forces and displace-

ments the sandwich’s top sheet are illustrated in

Figures 10 to 12. The plot of each sample will be created

with a particular colour, and symbol ~marks the point

with the maximal force in each test.

In the same kind of test, the resulting force and dis-

placement are not exactly the same between the sam-

ples. Beside common measurement errors, the most

important factor is the change at the samples’ bound-

aries. Because the naps are diagonally distributed

whilst the cutting lines are horizontal or vertical to

the panels’ borders, many naps are cut apart. Coming

from sample to sample, the cutting lines are not at the

same place, so the shapes of incomplete naps along

every boundary change as well (see Figure 9).

That makes the result fluctuate in the end. In the prep-

aration of the samples for each test, the differences of

the boundary geometry between the samples are

intended to examine how it will affect the result.

Figure 10–left reveals the results of P1-10 nap-core

sandwich samples in the compression test. They all

showed a typical behaviour under compression, but

the values of their force and displacement scattered a

bit. The maximum force changed between 1100 N to

around 1250 N, and the displacement at the buckling

altered between 0.44 mm and 0.51 mm. There are sev-

eral explanations for these differences. Firstly, the dif-

ferences in the boundary geometry make the results

altered. Secondly, the resin content of the samples is

not perfectly the same from sample to sample. Thirdly,

the curing process could not make a strictly uniform

effect on the entire big panel of the nap-core sandwich

– which would be divided into many separated samples.

Figure 10–right demonstrates the behaviour of P2-8

nap-core sandwich samples in compression was like

that of P1-10 nap-core sandwich samples. However,

the establishment period was much shorter, only

occurred when the displacement was smaller than

0.075 mm and the force was less than 100 N.

Although the displacement at the buckling is not so

Figure 7. Fixture schemes for the tests: Compression (left) and Shear (right).

Figure 8. Fixture schemes for the four-point bending test.

Ha et al. 373

different (0.34–0.36 mm for P2-8 nap-core sandwich,

and 0.44–0.51 mm for P1-10 nap-core sandwich), the

maximum force of P2-8 nap-core sandwich samples is

markedly higher (1750–2100 N for P2-8 nap-core sand-

wich, and 1100–1250 N for P1-10 nap-core sandwich).

Among the samples, the maximum force and the

displacement at buckling are pretty different, and the

reasons are very similar to those of P1-10 nap-

core sandwich.

As shown in Figure 11–left, the force and displace-

ment of the samples were not much scattered, and the

deformation did not contain a clear establishment

period. All of the samples deformed linearly as the

test started until the force went up to 1500 N and the

displacement was equal to 0.22 mm. Later, the samples

deformed nonlinearly. The force kept increasing to

more than 3000 N. When the displacement was over

0.85 mm, the shear buckling would happen suddenly

before the displacement reached 1.2 mm. Of five sam-

ples, two had the force reducing very slightly after the

shear buckling, and three had the force declining steep-

ly after the shear buckling. That was resulted from the

difference of the cohesion strength of the sandwich

samples. The phenomenon that the force plunged is

an indication of the delamination of the entire upper

sheet. Mostly, the weaker the cohesion strength is, the

sooner the delamination of the upper sheet happens. In

general, the delamination happens abruptly and there is

difficulty predicting its commencement precisely, but it

can be delayed by improving the quality of

the adhesive.

In Figure 11–right, the samples show the same

behaviour but much higher range of force. When the

displacement was less than 0.2 mm, the samples’ defor-

mations were nearly linear, and the forces increased

fast from 0 to 2600 N at least and 3400 N at most.

Afterwards, the samples behaved nonlinearly.

The force still went upward fast until the shear buck-

ling happened at a displacement between 0.6 mm and

0.8 mm. The maximum force (at the shear buckling)

changed somewhat from sample to sample. It ranged

from around 4750 N to around 5250 N.

As shown in Figure 12–left, the samples underwent a

quick nonlinear establishment period when the dis-

placement was less than 0.75 mm. Subsequently, they

behaved almost linearly until the initiation of the

damage (i.e., the local debonding of the upper sheet).

Based on the charts, it is viewed that five samples acted

not identically. Two samples buckled when the force

was around 75 N and the displacement increased over

6 mm for the one and 7.5 mm for the other. The third

sample worked linearly until the force got more than

90 N, and its damage occurred when the displacement

was about 6.5 mm (the blue chart). The remaining two

samples continued to work as the force went up above

105 N, and they only buckled when the displacement

Figure 9. Nap-core samples having the same dimensions but different boundaries.

Figure 10. Experimental results of the compression tests: P1-10 sandwich (left) and P2-8 sandwich (right).

374 Journal of Reinforced Plastics and Composites 38(8)

was more than 7.5 mm. The differences in the

four-point bending tests of the samples demonstrate

an obvious influence of the nap-core sandwich’s

boundary geometry.

Figure 12–right indicates that the samples of P2-8

nap-core sandwich performed very similarly as the

samples of P1-10 nap-core sandwich. The results, con-

sisting of the force and the displacement, also scatter to

some degree, but they are lower than those of P1-10

nap-core sandwich samples. Namely, the maximum

force is between 50 N and 70 N, and the displacement

at the debonding of the upper sheet is between 3.9 mm

and 5.5 mm. Once again, the influence of the sand-

wich’s boundary geometry is observed.

Through the experiments, buckling of the nap-core

and debonding of the upper skin can also be viewed

clearly. They are displayed by Figures 13 to 15.

Discussions

It can be noted that the sandwich acts almost like a

typical linear elastic material before the damages.

Moreover, all samples show good average material

strengths and moduli on density, which are shown in

Table 2.

There are a number of factors of the nap-core that

affect the obtained results of the tests on the sandwich

samples, i.e. the boundary geometry, the resin content,

Figure 11. Experimental results of the shear tests: P1-10 sandwich (left) and P2-8 sandwich (right).

Figure 12. Experimental results of the four-point bending tests: P1-10 sandwich (left) and P2-8 sandwich (right).

Figure 13. Buckling of the nap-core sandwich samples in the

compression test.

Figure 14. Shear buckling of the nap-core sandwich samples in

the shear test.

Ha et al. 375

and the curing condition. Of them, the boundary geom-

etry is the major one, and its influence is most consid-

erable in the four-point bending tests. To have a clearer

insight into the role of the nap-core in the sandwich’s

strength, the amount and aspect ratio of the nap-core

per area are taken and displayed in Tables 3 and 4, of

which R1 is the ratio of the mass of the nap-core to the

mass of the whole nap-core sandwich, and R2 is the

ratio of the top area of the nap-core to the base area of

the whole nap-core sandwich. It is noted that for each

sandwich sample, there is a trend that the bigger the

amount and the aspect ratio of the nap-core per area,

the higher the value of the maximal force. This is not

absolutely correct in every case but it can be seen an

appropriate ground to compare the maximum forces of

the nap-core sandwich samples in the same load-

ing status.

In the compression, the samples first underwent a

nonlinear interim period in which new contacts were

established since the nap-core is made with knitted

fabric. Subsequently, each nap of the nap-core provid-

ed the full resistance and the sandwich deformed

Figure 15. Local debonding of the nap-core sandwich’s upper

layer in the four-point bending.

Table 4. The mass ratio and the aspect ratio of the nap-core per area for P2-8 nap-core sandwich.

P2-8 nap-core sandwich Compression Shear Four-point bending

Sample no. 1 2 3 451234512345

0.264 0.269 0.271 0.272 0.274 0.217 0.219 0.220 0.222 0.223 0.208 0.211 0.214 0.216 0.219

0.344 0.348 0.352 0.359 0.365 0.344 0.348 0.352 0.359 0.367 0.340 0.344 0.352 0.359 0.367

Max force (N) 1762 1829 1865 1983 2088 4766 4854 4892 5013 5261 52.6 61.9 63.4 65.6 69.9

Table 3. The mass ratio and the aspect ratio of the nap-core per area for P1-10 nap-core sandwich.

P1-10 nap-core sandwich Compression Shear Four-point bending

Sample no. 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

0.492 0.494 0.498 0.500 0.503 0.440 0.441 0.443 0.445 0.451 0.413 0.434 0.437 0.438 0.442

0.284 0.288 0.290 0.293 0.295 0.289 0.290 0.293 0.294 0.298 0.281 0.286 0.292 0.295 0.299

Max force (N) 1109 1189 1197 1222 1252 3174 3187 3190 3206 3456 73.2 77.1 91.6 105.1 108.1

Table 2. Average densities and outcome values of the sandwich samples used for the experiment.

Nap-core sandwich type P1-10 P2-8

Test Parameters Value Standard deviation Value Standard deviation

Compression Elasticity modulus (MPa) 26.6 5.26 24.57 0.72

Compressive strength (MPa) 0.51 0.06 0.74 0.02

Sample density (kg/m

) 183.2 2.63 175.06 2.13

Specific strength (kN.m/kg) 2.78 4.23

Shear Shear modulus (MPa) 12.25 0.48 11.26 2.88

Shear strength (MPa) 0.32 0.01 0.43 0.11

Sample density (kg/m

) 182.3 1.30 174.10 2.50

Specific strength (kN.m/kg) 1.76 2.47

Four-point bending Flexural modulus (MPa) 1890.58 165.33 3241.82 273.59

Flexural strength (MPa) 7.32 0.94 6.06 0.53

Sample density (kg/m

) 181.9 4.04 174.68 1.24

Specific strength (kN.m/kg) 40.24 34.69

376 Journal of Reinforced Plastics and Composites 38(8)

linearly until the buckling happened, which caused a

quick downgrade of the force.

In the shear, there was not a clear period of estab-

lishment as that in the compression case, and the sand-

wich also performed linearly before that happened the

shear bucking. Interestingly, the force did not

descended responsively but kept unchanged for a

time. This phenomenon also occurred in shear test of

sandwiches with aluminium honeycomb, reviewed by

Franc¸ ois et al. in 2006,

and the reasons are not sim-

ilar. There was hardening character of the metal mate-

rial within the honeycomb core while there was yarn

jamming within the knitted fabric nap-core. Normally,

yarn jamming occurs when the fabric is extended to one

direction (either weft or warp); thus, the spacing

between the adjacent yarns in the other direction is

gradually minified; the yarns then get in contact and

hold one another better. In a shear case of the nap-core

sandwich, the extension of the nap-core’s knitted fabric

is not uniform, so there is also a local accruement of the

yarns in the fabric, which keeps the nap-core from a

collapse. In the end of the shear tests, the second

damage (i.e., the entire debonding of the top layer)

might happen.

In the four-point bending, the sandwich worked sim-

ilarly as itself in the shear test at first. The upper sheet

delaminated locally around the places where the

stresses are applied and that led to the plunge of

the force.

When compare the experimental samples of the two

nap-core types to one another, it is noticeable that the

sandwich of P2-8 symmetric nap-core has the higher

strength in the compression and shear tests while the

sandwich of P1-10 nap-core has the higher strength in

the four-point bending test.

One marked reason making P2-8 nap-core sandwich

has very good mechanical properties is its geometry.

Namely, the top diameter of its naps is rather small,

which is only 5.5 mm compared to 9.5 mm of the other

nap-core types. Therefore, the density of naps within

P2-8 nap-core is very thick, giving its sandwich high

strength. On the other hand, P1-10 nap-core possesses

good mechanical properties since it has a greater thick-

ness, heavier volume weight, and stronger fibre com-

pared to P2-8 nap-core.

Conclusion

The experimental results have demonstrated the essen-

tial attribute of the nap-core sandwich, which implies a

great possibility for numerical simulation of at least the

linear stage and initiation of the damage if not the

whole progression. Its linear elasticity is really different

to normal dry knitted fabric (without resin), which is

usually non-linear. Although the two nap-core types

have knitting patterns and they are different at many

other elements, the sandwich samples of both of them

behave mechanically in an identical way. If not

count the interim period in the compression tests,

the nap-core can be considered as a shell structure

which is much easier to be modelled with finite element

software.

In actual applications, the differences in the bound-

ary geometry of the nap-core sandwich samples are

inevitable, and their effect to the performance of the

samples may be considerable, especially in the bending

condition. Therefore, variety of boundary geometries

needs to be taken into account, and every design for

the use of nap-core sandwich should be computed on

the weakest case.

With fast development of the nap-core sandwich,

finding simulation methods for it are so necessary

because they will permit more cost-effective investiga-

tions on a wide range of the nap-core sandwiches, par-

ticularly when their parameters alter a great deal.

Furthermore, different applications may require differ-

ent parameters and specifications of the nap-core sand-

wich, and computational modelling is the most efficient

and quickest way to optimize designs of the sandwich

structures and predict their mechanical performances

in advance. This has been developed by the authors

and will be presented in a different paper.

Acknowledgements

The authors would like to thank The Institute of Mechanics

of TU Berlin, Fraunhofer Institute Pyco, and InnoMat

GmbH for constantly supporting this research.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with

respect to the research, authorship, and/or publication of

this article.

Funding

The author(s) received no financial support for the research,

authorship, and/or publication of this article.

ORCID iD

Giap X Ha http://orcid.org/0000-0003-3236-3968

References

1. Campbell FC. Introduction and uses of lightweight mate-

rials. Ohio, USA: ASM International, 2012.

2. Cantor B, Grant P and Johnston C. Automotive engineer-

ing: lightweight, functional, and novel materials. Florida,

USA: Taylor & Francis Group, 2008.

3. Ashby MF, Evans T, Fleck NA, et al. Metal foams: a

design guide. Oxford, UK: Elsevier, 2000.

Ha et al. 377

4. Valvano S and Carrera E. Multilayered plate elements

with node-dependent kinematics for the analysis of com-

posite and sandwich structures. Facta Universitatis-Series

Mechanical Engineering 2017; 15(1): 1–30.

5. Mallick PK (ed.). Materials, design and manufacturing for

lightweight vehicles. 1st ed. Cambridge, UK:

Elsevier, 2010.

6. Lightweight materials for cars and trucks. Department of

Energy, https://energy.gov/eere/vehicles/lightweight-

materials-cars-and-trucks (accessed 2 October 2017).

7. Courtney TH. Mechanical behaviour of materials: second

edition. Illinois, USA: Waveland Press, 2005.

8. Gerber N, et al. Symmetrical napcore and honeycomb

sandwich structures under impact load. SpringerLink,

https://link.springer.com/article/10.1007/s12221-016-

6271-8 (accessed 11 November 2017).

9. Gibson LJ and Ashby MF. Cellular solids: structure

and properties. Cambridge, UK: Cambridge University

Press, 1999.

10. Wadley HNG, Fleck NA and Evans AG. Fabrication and

structural performance of periodic cellular metal sandwich

structures. Compos Sci Technol 2003; 63: 2331–2343.

11. Gerber N. Noppenwaben als neue Kernwerkstoffe fu

¨r

Leichtbauanwendungen (Napcores as new core material

for lightweight applications). PhD Thesis, Brandenburg

University of Technology, Brandenburg, Germany, 2016.

12. Nils G, Andreas B, Christian D, et al. Progress on the

development and automated production of new core mate-

rials. Berlin, Germany: Fraunhofer Research

Institute, 2017.

13. Ha GX and Zehn MW. Experimental study and finite

element simulation of novel nap-core sandwich compos-

ite. Compos B Eng 2019; 158: 117–130.

14. Bernaschek A, Chowdhury Y and Kahle O. Nap core –

core material for sandwich construction. In: Proceedings

of the 2nd International Conference on Thermosets, Berlin,

Germany, 21-23 September 2011.

15. Monica B, Peter F and Christoph U. Sandwiches with

napcore. Kunststoffe International. Munich, Germany:

Carl Hanser Verlag, 2006, pp.95-97.

16. Franc¸ ois C, Vikram SD and Norman AF. The shear

response of metallic square honeycombs. J Mech Mater

Struct 2006; 1: 208.

378 Journal of Reinforced Plastics and Composites 38(8)