d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 12341244

Available online at www.sciencedirect.com

ScienceDirect
journal homepage: www.intl.elsevierhealth.com/journals/dema

The effect of ber aspect ratio and volume loading

on the exural properties of owable dental
composite
Paul Shouha, Michael Swain, Ayman Ellakwa
Biomaterials Science, Discipline of Biomaterials, Faculty of Dentistry, The University of Sydney, Sydney Dental
Hospital, 2 Chalmers St., Surry Hills, Sydney, NSW 2010, Australia

a r t i c l e

i n f o

a b s t r a c t

Article history:

Objective. To evaluate the efcacy on exural properties of owable dental resin composite

Received 25 September 2013

reinforced with short glass ber of different aspect ratios (ARs) and volume percent load-

Received in revised form

ings. It is hypothesized that with the addition of randomly oriented bers it is possible to

2 April 2014

signicantly improve exural strength and modulus while maintaining owability.

Accepted 8 August 2014

Methods. Ten groups of samples with varying glass ber volume loads (0, 5%, 10%, 20%, 40%
and 60%) and three different ARs (5.2, 68 and 640) were tested in three point bending to
fracture according to ISO 4049. A owable resin composite was used as the control and also

Keywords:

as the lled resin composite that was subsequently reinforced with bers. Load deection

Flexural

results were used to calculate exural strength and exural modulus. SEM images were

Fracture

used to determine the mode(s) of failure, to describe surface features of reinforcement and

Strength

were correlated with force displacement graphs. All results were statistically analysed using

Modulus

ANOVA followed by post hoc Tukeys test. Level of signicance was set at 0.05.

Dental composite

Results. When compared to the sculptable control (68.6 vol% ller loaded) results for ex-

Flowable

ural strength varied from a mean reduction of 42% (p > 0.05) for the low AR group to an

Glass ber

increase of 77% (p < 0.001) for the high AR samples. Flexural modulus results varied from a

Aspect ratio

low of 6.6 [0.67] GPa for the non reinforced spatulated control to 20.3 [1.31] GPa (p < 0.001) for
the 60% loaded low AR group. The low ber loaded mid AR group was still owable with 49%
total loading (5% ber/44% ller) but gave strength values (181.2 [33.5] MPa) 30% higher than
the sculptable control (p > 0.05) and comparable modulus.
Signicance. This study shows that short and very short glass bers can signicantly reinforce
owable dental composite. The bers aspect ratio was shown to be more important than volume loading for exural strength. It appears possible to produce a light cured short glass ber
reinforced owable material with superior exural properties compared to conventional
universal composites.
2014 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

Corresponding author: Associate professor, Faculty of Dentistry, The University of Sydney, Westmead Oral Health Centre, Level 1,
Westmead Hospital, Darcy Road, Sydney, NSW 2145 Australia. Tel.: +61 298457161; fax: +61 296334759.
E-mail address: ayman.ellakwa@sydney.edu.au (A. Ellakwa) .

http://dx.doi.org/10.1016/j.dental.2014.08.363
0109-5641/ 2014 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 12341244

1.

Introduction

Whether driven by public esthetic demands, concerns over

amalgams toxicity or trends toward conservative management of dentitions, dental composites have become the
material most often employed in the restoration of teeth.
These materials are composed of varying amounts of inorganic radio opaque particulate llers dispersed in a relatively
weak organic resin matrix. Bi-functional silane compounds
are used to create the interphase between the two. These
materials are manipulated directly in the mouth and are set
with curing lights of approximately 470 nm wavelength. Their
use in stress bearing areas of posterior teeth is becoming an
acceptable option but brittleness and a tendency to edge fracture remain important limitations [1].
There has been a long and successful history of high
performance ber reinforcement of industrial composites.
This cannot be said of dental composites. Although relatively
uncommon clinically, the introduction of bers have been
studied in xed and removable dental prostheses [2,3]. These
ber systems though are usually continuous and are indirectly
manufactured [46]. Compared with direct procedures these
restoratives require more aggressive tooth preparation. Other
limiting factors include signicant laboratory fees and the
increased probability of handling and processing errors. Chair
side reinforcement with continuous ber composite is still
less common. This direct procedure involves combining regular dental composite with sections cut from straight or woven
ultra high molecular weight polyethylene bers. Not only has
ber reinforcement been shown to strengthen and toughen
dental composite but direct reinforcing was also found to
strengthen restored teeth [7]. Additionally, the fracture type
changes from a catastrophic brittle pattern to a more graceful
controlled one.
Least common are dental composites reinforced with randomly oriented short bers. These materials claim higher
stiffness, fracture toughness and exural strength. For example: 16 GPa, 3.2 MN/m3/2 and 176 MPa respectively for a 71%
volume lled, 20800 m long E glass rod material [8,9]. They
also claim very low polymerization shrinkage [10] but have
been found to exhibit poor wear properties due to plucking of the relatively large surface exposed bers and llers
[11].
Variables such as ber position and orientation have
been studied [1214]. Using chopped glass bers in denture polymers Karacaer [15] found signicant correlation
between ber volume with both exural modulus and impact
strength. Callaghan [16] found that increasing the length
of the bers increased the wear resistance of reinforced
specimens.
The intent of this study was to evaluate the effect of
different ber aspect ratios and ber volume loads on the
exural strength and modulus of low viscosity or owable
dental composites. Fiber aspect ratio (AR) is simply the relationship of length to diameter of a ber given as a number.
Some studies have quantied AR as a variable for mechanical properties of dental materials [1618] but overall there
exists very little research in this eld and none that we know
of with regards to owable dental materials. The hypothesis

1235

evaluated was that a owable dental composite may be signicantly reinforced by incorporating randomly distributed
short and very short glass bers. Further it was hypothesized
that, the amount of ber required to increase the exural
behavior is small thereby retaining low viscosity and owability.

2.

Materials and methods

Ten groups (AJ) with ten samples per group were formed.
Three control groups A, B and C were not reinforced while
groups DJ formed the ber reinforced samples (Table 1). Group
A consisted of a universal high viscosity or sculptable dental composite (Beautil II, Shofu, Kyoto, Japan) while groups B
and C consisted of a low viscosity owable composite (Beautil F03 Flow Plus). Groups DF were formed by combining
this owable material (F03 Flow Plus) with low (5.2) aspect
ratio bers of different volume load fractions (20, 40 and 60%).
Groups G, H and I combined F03 Flow Plus with mid (68)
aspect ratio bers of three different loads (5, 10 and 20%)
while group J utilized high aspect ratio (640) bers with 10%
loading.

2.1.

Fibers

Glass bers with three different aspect ratios (ARs) were used
(Fig. 1). Low AR bers presented as unsized milled E glass powder from Owens Corning, Ohio, USA. SEM analysis revealed an
average length of 89 62 m (range 14300, n = 400), an average
diameter of 17 m and an AR of 5.2 3.65. They were used for
three groups (DF). High AR glass bers presented as S glass (S2 Glass) chopped strands from AGY, South Carolina, USA and
consisted of numerous 10 m diameter parallel laments cut
into short 6.4 mm lengths. These bers have an AR of 640 without deviation and were used for group J. These S glass strands
were also cut (average length 679, range 602400, n = 400) to
create the third AR of 68 43.5 used in 3 groups (GI).
Both E glass and S glass bers were used as the availability of very short cut or milled glass bers is very limited.
The prex E and S stands for electrical grade and structural grade. Their physical properties are as follows: tensile
strengths 3.4 GPa vs. 4.9 GPa, moduli 72 GPa vs. 87 GPa and
densities of 2.58 g/cc vs. 2.46 g/cc respectively. Both bers are
predominantly aluminosilicate glasses. E glass is high in calcium while S glass is high in magnesium with no boron
[19]. Given the similarity in composition, physical properties
and surface chemistry it was decided to use milled E glass
as a low AR baseline. This material is readily available and
is used by manufacturers of packable DRCs. The S glass
bers were used uncut in the higher AR sample group. The
major difference between the two glasses was thus considered to be diameter and a sufcient spread of aspect ratios was
achieved.
All bers were rst etched in a 4% potassium
dichromate(IV)(K2 Cr2 O7 )/89% sulphuric acid (H2 SO4 ) aqueous
solution (Australian Scientic, NSW, Australia) for 4 h then
thoroughly washed (until pH returned to 7) in deionized
water followed by rinsing in ethanol and drying in an oven
at 50 C for 8 h. Fibers were then immersed in a hydrolyzed

1236

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 12341244

Table 1 Ten groups (n = 10): three controls (AC); three low aspect ratio (DF); three mid AR (GI) and one high AR (J)
groups. Mean values and one standard deviation (brackets) are given. Note that only group G remained owable after
ber incorporation.  ef , effective stress yielding; Eef , effective modulus.

Group

Composion

Resin

Flowability

Flexural Strength
mean

(SD)

MPa

ef

Flexural Modulus
mean (SD)

GPa

Eef

Beautil II universal giomer composite

31.4

Sculptable

139.8

(10.6)

12.2

(0.74)

Beautil Flow Plus F03 owable giom er syringed

53.7

Low ow

126.3

(11.8)

7.6

(1.0)

Beautil Flow Plus F03 owable giom er spatulated

53.7

Low ow

103.5

(18.3)

6.6

(0.67)

F03 + 20% low AR (5.2 ) E Glass

43

Paste

95.2

(15.9)

62

11

(1.11)

28.6

F03 + 40% low AR (5.2 ) E Glass

32.2

Paste

94.1

(11.5)

80

17

(2.02)

32.6

F03 + 60% low AR (5.2 ) E Glass

21.5

Packable

81.2

(19.2)

66

20.3

(1.31)

29.4

F03 + 5% mid AR (68) S Glass

51

Flowable

181.2

(33.5)

1658

12.1

(0.94)

116.6

F03 + 10% mid AR (68) S Glass

48.3

Paste

170.4

(28.2)

772

13.8

(2.39)

78.6

F03 + 20% mid AR (68) S Glass

43

Paste

210.7

(48.7)

639.5

13.2

(2.85)

39.6

F03 + 10% high AR (640) S Glass

48.3

Paste

247.7

(75.2)

1545

13.3

(1.34)

73.6

Flowable composites are indicated in red in Flowability column. Blue and brown values relate to effective stress yielding and modulus described
on lines 364402.

silane solution consisting of 2% 3-(Trimethyoxysilyl) propyl

methacrylate silane (Z-6030 SigmaAldrich, MO, USA) in
93% ethanol and 5% deionized water. This solution was
titrated to pH 5 with acetic acid. The bers in solution
were stirred for 2 min every 5 min for a total of 44 min
before being rinsed several times in deionized water. In
order to remove any physisorbed and some of the weakly
chemisorbed layers the bers were then conditioned in 100 C
deionized water for 10 min and then dried at 50 C for 24 h
[20,21].

2.2.

Dental resin composite

The dental composite used in this study Beautil is classied as a giomer and is composed of BisGMA/TEGDMA
resin and pre-reacted glass-ionomer (S-PRG) uoroaluminosilicate glass ller particles ranging in size from 0.01 to 4.0
(ave. 0.8) m. Manufacturers of these materials claim uoride
release and recharge capabilities while maintaining superior
physical properties to poly acid modied resin composites
(compomers) [22].

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 12341244

1237

Fig. 1 E glass (a) and S glass (b) as received from manufacturer. Cut S glass bers (c and d) showing random inter
penetrating networks.

This material is manufactured with ve viscosities depending on ller ratio ranging from 68.6% volume (83.3% by weight)
loaded sculptable Beautil II (group A) to the very low viscosity Beautil Flowable F10 (<47% by volume). Beautil Flow
Plus F03 (shade A2 lot number 111004) with 46.3% by volume
(67.3% weight) ller loading and mid viscosity was chosen as
the main control (group B) and also as the resin composite to
be reinforced with glass ber as it exhibited good strength at
low viscosity.

2.3.

Sample preparation

All samples were prepared according to ISO 4049. A split

brass mold (2 mm 2 mm 25 mm) with glass top and base
was used. The samples were cured using a Flash Max P3 460
(CMS Dental ApS, Copenhagen, Denmark) curing light according to manufacturers instruction (3 s) with six overlapping
cures from both top and bottom slides. A standard light output meter (range 02000 mW) showed light output to be above
2 W. The packable control was dispensed onto a paper mixing
pad before packing into the mold. The owable composite was
dispensed directly from the manufacturers syringe dispensing tip (internal = 600 m) into the split metal mold while
all reinforced samples were mixed individually on the paper
mixing pad with a metal spatula for 20 s prior to packing
into the mold. To assess the effect of mixing on the samples

another control group C was prepared by rst dispensing the

owable resin composite F03 onto the pad and then spatulating for 10 s before loading into the mold.
The weight of ber needed for each sample was found by
using the densities of the two bers used and also the DRC
(1.80 g/cc). Density () equals weight in grams divided by volume in cubic centimeters. Therefore,
Volume =

Weight
Density

and since,
%Vf =

Vf
Vf + Vr

where Vf + Vr = Total Sample Volume

Therefore,
%Vf =

Wf /f
Total Sample Volume

Fibers for each test sample were thus weighed according to

the following formula:
Wf = %Vf f Total Sample Volume (TSV)
(where Wf is the weight of ber in grams, %Vf is the percentage volume ber in sample and f is the density of ber
in grams/cubic centimeter, TSV = 2 mm 2 mm 25 mm =
0.1 cm3 ) before being mixed into the individually weighed

1238

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 12341244

owable composite. Samples were stored in deionized water

at 37 C for 24 h prior to testing.

3.

Testing

The specimens were tested in three-point bending loaded

at 1 mm/min on a 20 mm span using a universal testing
machine (Shimadzu Autograph AG-50 kNE, Shimadzu, Japan).
The elastic moduli were obtained from the Shimadzu program.
Flexural strength (FS) was determined from the following
equation:
FS =

3PL
2wt2

where P is the peak load in Newtons; L is the distance in mm

between support struts; w is the width of specimen in mm;
and t is the thickness of specimen in mm.
All results were statistically analyzed using ANOVA followed by post hoc Tukeys test. Level of signicance was set
at 0.05.

4.

Results

Mean and effective (explained later in discussion) values for

exural strength (FS) and exural modulus (FM) with one
standard deviation are given in Table 1 and shown graphically
in Fig. 2.

4.1.

Flexural strength

The packable control group A and owable control groups B

and C gave means of 139.8 (10.6) MPa, 126.3 (11.8) MPa and
103.5 (18.3) MPa respectively. All low aspect ratio samples were
weaker than controls though not signicant at p < 0.05. The
lowest mean strength value was seen with 60% low aspect
ber group F with an average of 81.2 (19.2) MPa, a 21% decrease
from the spatulated control group C.
All mid and high aspect ratio groups showed signicant
strength augmentation to control group C at p < 0.05. The
highest mean was seen with 10% loaded high aspect random ber group J with 247.7 (75.2) MPa, a 77%, 96% and 139%
improvement over universal control A, syringed control B and
spatulated control C respectively (all p < 0.001).

4.2.

Fig. 2 Flexural strength in MPa (a) and elastic modulus in

GPa (b). Groups within horizontal lines not signicant at
p < 0.05. Vertical lines show 1 SD.

Flexural modulus

The packable control group A and owable control groups B

and C gave means of 12.2 GPa (0.74), 7.6 GPa (1.0) and 6.6 GPa
(0.67) respectively. As expected all reinforced groups tested
were signicantly stiffer at p < 0.001 than the owable controls
B and C while the 40% and 60% low aspect ratio samples were
also signicantly stiffer at p < 0.001 than the packable control
A with the low strength 60% volume loaded low aspect group
E achieving a 66% improvement with 20.3 GPa (1.31).
Interestingly, one of the reinforced groups (G) retained relative owability yet outperformed the sculptable more highly
packed control composite (group A).

4.3.
Scanning electron microscopy and force
displacement graphs
Scanning electron microscopy analysis (Fig. 3) showed surface
and subsurface voids and preparation aws in all samples. The
large 400 m void in Fig. 3a resulted in a very low performance
for this sample. As expected, the spatulated control showed
much greater incidence of small and large air voids (see Figs. 3a
and b white arrows).
Scanning electron microscopy of all the controls and low
aspect ratio samples showed relatively at fracture surface
topography (see Fig. 3a and b) and correlated with brittle force
displacement graphs (Fig. 5 blue line). Photo micrographs of
10% and 20% mid aspect and all high aspect samples showed
very rough proles (Fig. 3d) and correlated with tougher force
displacement graphs (Fig. 5 green line). These samples demonstrated the presence of a deection hardening stage II which
corresponds to the onset and then saturation of ber engagement and matrix micro cracking. This stage does not exist
when ber volume is less than the critical-bending volume
and this was seen with all of the 5% mid aspect ratio samples which showed rough photo micrograph proles (Fig. 3c)
but still exhibited brittle performance in force displacement
proles.
The presence of a stage III tail is typical of further ber
debonding and pullout with associated energy expenditure.
All of the reinforced samples showed steeper inclines indicating higher modulus. Also, the end of the elastic phases

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 12341244

1239

Fig. 3 (a) Fractured surface of syringed owable control sample showing very large (400 m) surface aw (black arrow); (b)
sample from spatulated owable control group showing greater number of small (<40 m) and large (200 m) internal voids
(white arrows). Note the relatively at fracture surface topography. (c) Sample from 5% volume loaded mid AR group
showing very rough fracture surface and ber bridging (red arrows). (d) 10% volume loaded high AR overall best performing
sample. Even though this sample performed very well there was still the presence of voids and smooth ber surfaces. Note
long ber perpendicular to the long axis and near the surface acting as a stress concentrator (arrow). Also note the non
random distribution of bers. (For interpretation of the references to color in this gure legend, the reader is referred to the
web version of the article.)

occurred at much higher loads in the reinforced samples (see

Fig. 5 with 81 N vs. 35 N).

aramid bers are yellow and carbon bers and nano tubes are
black.

5.2.

5.

Discussion

5.1.

Use of glass bers

Glass bers were used in this study because they are very conducive to silane bonding without the need for extra treatment.
Crystalline ceramic whiskers utilized by Xu et al. [6] required
surface fusion with nano silica particles at 800 C to achieve
effective stress transfer. Organic polyethylene bers require
plasma etching prior to silanization and still exhibit relatively
low bond strengths. The bers used in this study are also safe
to handle as they are signicantly greater than the 3 m ber
diameter considered dangerous in lung tissue and have low
biopersistence when compared to refractory ceramic bers
[23,24]. Glass bers have a refractive index much closer to
dental resins and allow for light polymerization as opposed
to crystalline whiskers, carbon bers and carbon nano bers.
The comparable refractive index also makes glass ber composites esthetic whereas whiskers at present are very opaque,

Fiber aspect ratio

It has been shown that discontinuous ber systems rarely

exceed 70% of the strength values achieved with continuous systems [25]. However chopped ber composites can
reach strength values near that of continuous systems so
long as ber length or aspect ratio exceeds a critical value.
Alander et al. [26] reported exural strength of continuous/unidirectional ber reinforced dental resin composite to
be between 800 and 1200 MPa. According to Chen [25] therefore, with an optimized discontinuous system it may be
possible to achieve up to 560840 MPa, i.e. 70% of exural
strength.
The critical ber length lcrit necessary to maximize stress
transference to the bers is related to ber diameter d, ber
tensile strength  f and ber matrix bond strength  by the
following equation [17]:

lcrit =

f d
2

1240

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 12341244

Fiber fragmentation tests have shown the critical length of

E-glass bers in Bis-GMA varies between 0.5 and 1.6 mm [27].
Assuming a diameter of 17 m and given AR = length/diameter
this equates to aspect ratio range of 3094. Kardos [28] found
the critical AR for modulus occurs at about 100.
Garoushi et al. [27] studied experimental short glass
ber reinforced DRCs observing signicant improvements in
mechanical properties (dry exural strength values of 210 MPa
compared with 111 MPa for a universal hybrid composite control) but the 3 mm E glass bers used were relatively long with
limited clinical use. A recent study by Garoushi et al. [29] utilizing shorter E glass bers 1.32.0 mm also found signicant
improvements in exural strength and exural modulus and
is in agreement with results found in this study.
Other work has focused on ceramic whiskers. Xu et al.
[6] using randomly distributed 0.5 m diameter silicon nitride
whiskers (aspect ratio 10) silane treated and surface fused
nano (40 nm) silica particles at 55:45 mass ratio to resin was
able to obtain exural strength and modulus values of 196 MPa
and 7 GPa respectively.
The results of this study support the hypothesis that very
short bers less than 1 mm in length can still have a signicant effect on exural properties and together with low
volume loading may produce an easy to use owable material.
Flowability allows for extrusion through a nozzle, simplifying placement and overcoming difculties, namely technique
sensitivity, time and expense [30].
The results also show that the use of bers with too low an
aspect ratio can actually weaken the composite as ber ends
create discontinuity in the matrix and act as stress concentrators or aws [31]. Most of the mid and high aspect ratio
reinforced samples showed marked and signicant increases
to strength and modulus (Tables 1 and 2) since the effect of
the ber ends decrease as the ber length increases. A doubling of exural strength was seen with the high aspect ratio
group J. Overall the relationship between exural strength and
aspect ratio within the range studied was positive and logarithmic linear (see Fig. 4a). Here, the ber aspect ratio range
for both high exural strength and workability was found to
be between 50 and 500 and is in agreement with Kardos [28]
who found that for strength, plateau values are attained with
an aspect ratio of about 500.
On comparing the 10% loaded groups (H and J), aspect ratio
was not a signicant variable with regards to modulus as can
be seen with the mid aspect group H at 13.8 GPa vs. 13.3 GPa of
the high aspect group J. But increasing the aspect ratio from 5.2
to 68 resulted in a 20% increase in modulus for the 20% loaded
groups D and I possibly indicating a critical aspect ratio (see
Fig. 4b). Beyond this critical value, volume loading fraction is
less relevant to modulus. Only 5% loading of the mid aspect
bers (group G) was able to achieve a modulus greater than
the 20% low aspect group D and comparable to the 68.9% lled
packable control group A. So with regards to modulus it may be
surmised that the critical aspect ratio is somewhere between
5.2 and 68.
With continuous bers a rule-of-mixture equation exists
such that:

c = f Vf + m Vm

or

Ec = Ef Vf + Em Vm

a
3

Fig. 4 The relationship between exural strength and

aspect ratio within the range studied was positive and
logarithmic linear (above). Modulus appears to be relatively
independent from volume loading beyond an
undetermined cut off aspect ratio (below).

where  c is the ber composite yield strength;  f is the ber

yield strength;  m is the matrix yield strength; E is the elastic modulus and V is the volume [32]. In order to quantify the
efciency of different ber aspect ratios on outcome measures
a modication of this equation can be used. Where cut ber
length is greater than 15lcrit the bers are termed continuous
and aligned. But, bers used in this study were very likely less
than 15lcrit especially given the less than ideal ber matrix
bond strength (see smooth ber surfaces in scanning electron
micrographs) and therefore largely discontinuous and random
in three dimension. Here, a ber efciency parameter K = 0.2,
is used [32] such that:
c = Kf Vf + m Vm

and

Ec = KEf Vf + Em Vm

However, most micrographs of the fractured surfaces

exhibit some aligning of bers (see Fig. 3ce) and this
may increase the K efciency factor (possibly to somewhere
between 20% and 60%).
But less than ideal ber matrix interface bond strength
 means that not all of the load delivered to the matrix is

1241

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 12341244

Table 2 Table showing percent differences (red, negative; black, positive) in exural strength and modulus with control
groups AC. *p < 0.05, **p < 0.001.

Group

AR

Fibre vol
%

Filler
bre
total vol
%

Sculptable control group A

Flowable control group B

Spatulated control group C

FS %

FS %

E%

FS %

11

** 61

35

** 85

22

15

E%

E%

68.6

46.3

13

** 38

46.3

26

** 46

18

5.2

20

57.0

32

10

25

* 45

** 67

5.2

40

67.8

33

** 40

25

** 124

** 158

5.2

60

78.5

42

** 66

36

** 167

21

** 208

68

49.0

30

43

** 59

** 75

** 83

68

10

51.7

22

13

35

** 81

* 65

** 109

68

20

57.0

** 51

** 67

** 74

** 104

** 100

640

10

51.7

** 77

** 96

** 75

** 139

** 102

transferred to the stronger ber and this has a negative effect

on K.If,
ef = Kc
where  ef is the effective stress fracture of short bers then the
efciency of bers with different dimensions can be measured
given the same conditions exist for surface treatment and 
[33].
ef =

c (1 Vf )m
Vf

Fracture strength and elastic modulus for the spatulated

control group C was used for the determination of  m and
Em . The different ber efciencies for strength and modulus
are seen in Table 1. It can be seen that for exural strength
the greatest efciency is seen with low load mid aspect and
mid load high aspect systems. With modulus the greatest
efciency was seen with the low load mid aspect system. Interestingly of all of the reinforced systems this was the only group
that could be considered owable.

5.3.

Fiber volume loading

The effect of varying the ber loading on strength and modulus of the owable composite was investigated with the 20%,
40% and 60% low aspect ratio samples (groups DF) and 5%,
10% and 20% mid aspect samples (groups GI). Note that
although containing 22.3% less ller, the owable control B
was only 13% weaker but 38% more exible than the packable
control A.
With regards to strength ber loading was not signicant
within the parameters studied in Fig. 2a showing a weakening
effect in the low aspect ratio samples. Also the low ber loaded

13

(5%) mid aspect ratio group G with only 49% combined ller
and ber loading showed a 75% (p < 0.001); 43% (p < 0.05) and a
30% (p > 0.05) improvement over the spatulated (46.3% lled),
syringed (46.3% lled) packable (68.6% lled) control groups C,
B and A. In fact, although while not signicant at p < 0.05 the
60% low aspect ratio group F showed a 36% drop over the owable control group B and a 42% drop over the packable control
group A indicating an increase in the negative inuence of
low aspect bers, i.e. matrix discontinuity and stress concentration. This effect may also be at least partly due to a greater
potential for air entrapment.
The 20% loaded mid aspect ratio group I showed a significant 67% and 51% increase over the owable and packable
controls respectively. Groups I and D with equal volumes of
bers (20%) showed marked differences in the effect of aspect
ratio with the longer bers resulting in a material twice as
strong (121% p < 0.001). Brandt [34] found that reinforcement
is proportional to ber volume but it is important to identify
a negative reinforcing effect past a certain volume. Garoushi
et al. [27] found that random short glass bers with less than
22% volume loading had a signicant strengthening effect
on dimethacrylate-polymethylmethacrylate resin based composite.
With regard exural modulus ber volume loading was proportional to modulus in the low aspect ratio systems only
(Fig. 2b). Increasing mid aspect ratio ber loading from 5% to
20% did not improve modulus beyond the 1214 GPa range possibly indicating a modulus loading dependence somewhere
above 20%. The 20% loaded mid aspect group I was 20% stiffer
than the 20% low aspect group D indicating a mild inuence of
aspect ratio on modulus. As expected signicant increases to
exural modulus was seen with all aspect ratios even at low
volume loads compared to the owable 46.3% lled control

1242

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 12341244

Fig. 5 Force displacement/stress strain curve overlays of the best performing syringed owable control (group B) sample
(blue) and best reinforced sample (from 10% high AR group J) (green). (For interpretation of the references to color in this
gure legend, the reader is referred to the web version of the article.)

group B. The addition of only 5% mid aspect ratio bers (group

G) to the owable composite resulted in an equal performance
to the 68.6% loaded largely spherical submicron particle control group A.

5.4.

Fracture behavior

The fracture toughness of a material is a measure of how

well that material hinders the progress of a crack or aw
under load. Fibers impede the extension of a crack and
develop interlocking bridges (see red arrows in Fig. 3c)
behind the progressing crack dissipating energy by ber
pullout resulting in graceful rather than catastrophic failure.
As expected the force displacement charts (summaries in
Fig. 5) showed elastic (I) and catastrophic fracture behavior in
the non reinforced controls. With the introduction of bers
the behavior changes to include an inelastic region (II) with
the onset and then saturation of ber engagement and matrix
multiple micro cracking also termed strain or deection hardening. Deection hardening stage does not exist when ber
volume is less than the critical-bending volume [34]. With
ber debonding and pullout further energy is expended (III)
and bers are seen to bridge the advancing macro crack (softening branch) prior to complete failure. Regions II and III were
not seen with most of the low aspect bers whatever the volume loading and was also not seen in most of the low volume
mid aspect samples indicating the predominance of brittle
resin composite properties.

5.5.

Mixing

The overall greater concentration of air pockets in the form

of small and large voids may be responsible for the spatulated control (group C) being 18% weaker and 13% more
exible (p > 0.05) than syringed control (group B) suggesting
that improvements in exural properties may be expected
with void minimization in the mechanical manufacturing of a
reinforced material. In addition it may be expected that ber
surfaces further entrap air so minimizing the volume loading
may help optimize properties.
Overall, the addition of randomly oriented very short (less
than 1 mm) bers led to signicant changes to both strength
and modulus of the owable composite although only the
5% loaded mid aspect ratio group retained owability. This
material had 73% the maximum exural strength of the best
performing test group but with bers 11% of the length and
50% of the volume. It also outperformed the sculptable universal control in strength by 30% (p > 0.05) while containing
62% more resin (51% vs. 31.4%).

6.

Limitations and future direction

The mixing of the mid and high aspect ratio bers resulted
in some clumping and paralleling of bers as can be seen in
Fig. 3c and d. The dispensing of a ber reinforced resin composite through a nozzle may show a similar result. Further
work may look at the inuence of ber aspect ratio, ber volume fraction and nozzle diameter in delivering a randomly
distributed ber system. In this study, viscosity was assessed

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 12341244

visually. Future work will involve quantitative measurement

of owability. This study has shown the signicant impact of
surface and other imperfections and inclusions on strength
properties of dental composites. Therefore, future work may
also look at surface manipulation and treatments.

7.

Clinical signicance

The aim of any material research is to improve accepted bench

top physical properties and practical clinical durability. To our
knowledge there are no dental resin composites with exural
strength values above 200 MPa (max. 177 MPa for 82% vol. nano
hybrid). Given dentin exural strength values of 260 MPa being
reported it would seem appropriate to mimic this material
property as close as possible if stress strain disparity is to be
avoided. The addition of bers have been shown to reliably
increase exural modulus. Given dentin exural modulus values reported average 17 GPa it may be advantageous to mimic
this parameter as well. It may be argued that while fracture
toughness is a measure of resistance to the catasrophic propagation of aws it more a materials strength that inuences
initiation. Thus bers may improve damage incidence as well
as tolerance.

8.

Conclusions

Within the parameters studied the ber aspect ratios of 68

and 640 resulted in signicant (p < 0.05 and p < 0.001) improvements to the exural strength of owable controls even with
relatively small volume load. All low aspect ratio (5.2) samples
though had a negative effect and ber volume loading was not
signicant for strength.
As expected with regards to exural modulus the presence of any volume of stiffer ber led to a signicant increase
(p < 0.05) in stiffness even if it weakened the composite (low
aspect ratio). Fiber aspect ratio was not signicant for modulus.

Acknowledgements
Ken Tyler for his time and expertise at the Biomaterials Laboratory, Sydney Dental Hospital.
The Australian Center for Microscopy and Microanalysis
(University of Sydney), in particular, Drs Ian Kaplin and Patrick
Trimby.
Shofu (Kyoto, Japan) and Horsley Dental (Australia) for their
supply of owable composite.
AGY (South Carolina, USA) for supply of S-2 glass bers.
Owens Corning (Ohio, USA) for E glass bers.

references

[1] Ilie N, Hickel R. Resin composite restorative materials. Aust

Dent J 2011;56(1 Suppl.):5966.
[2] Basaran EG, Ayna E, Vallittu PK, Lassila LVJ. Load bearing
capacity of ber-reinforced and unreinforced composite
resin CAD/CAM-fabricated xed dental prostheses. J
Prosthet Dent 2013;109:8894.

1243

[3] Narva KK, Lassila VL, Vallittu PK. The static strength and
modulus of ber reinforced denture base polymer. Dent
Mater 2005;21:4218.
[4] Al-Darwish M, Hurley RK, Drummond JL. Flexure strength
evaluation of a laboratory-processed ber-reinforced
composite resin. J Prosthet Dent 2007;97:26670.
[5] Nakamura T, Waki T, Kinuta S, Tanaka H. Strength and
elastic modulus of ber-reinforced composites used for
fabricating FPDs. Int J Prosthodont 2003;16:
54953.
[6] Xu HHK, Schumacher GE, Eichmiller FC, Peterson RC,
Antonucci JM, Mueller HJ. Continuous-ber preform
reinforcement of dental resin composite restorations. Dent
Mater 2003;19:52330.
[7] Belli S, Cobankara FK, Eraslan O, Eskitascioglu G, Karbhari V.
The effect of ber insertion on fracture resistance of
endodontically treated molars with MOD cavity and
reattached fractured lingual cusps. J Biomed Mater Res B:
Appl Biomater 2006;79B:3541.
[8] Biodental Technologies Internal Data. BDT, Sydney,
Australia.
[9] Zakaria MR, Najim WA. A comparison of fracture strength of
posterior composite and ceramic inlay materials. J Bagh
College Dent 2009;21:1822.
[10] Stavridakis MM, Lutz F, Johnston WM, Krejci I. Linear
displacement and force induced by polymerization
shrinkage of resin-based restorative materials. Am J Dent
2003;16:4318.
[11] Manhart J, Kunzelmann KH, Chen HY, Hickel R. Mechanical
properties of new composite restorative material. J Biomed
Mater Res Appl Biomater 2000;53:35361.
[12] Ellakwa A, Shortall A, Shehata MK, Marquis PM. The
inuence of ber placement and position on the efciency
of reinforcement of bre reinforced composite bridgework. J
Oral Rehabil 2001;28:78591.
[13] Karbhari VM, Strassler H. Effect of ber architecture on
exural characteristics and fracture of ber-reinforced
dental composites. Dent Mater 2007;23:9608.
[14] Dyer S, Lassila L, Jokinen M, Vallittu P. Effect of
cross-sectional design on the modulus of elasticity and
toughness of ber-reinforced composite materials. J Prosthet
Dent 2005;94:21926.
[15] Karacaer O, Tezvergil L, Vallittu P. The effect of length and
concentration of glass bers on the mechanical properties of
an injection and a compression moulded denture base
polymer. J Prosthet Dent 2003;90:38593.
[16] Callaghan DJ, Vaziri A, Nayeb-Hashemi H. Effect of ber
volume fraction and length on the wear characteristics of
glass ber-reinforced dental composites. Dent Mater
2006;22:8493.
[17] Petersen RC. Discontinuous ber-reinforced composite
above critical length. J Dent Res 2005;84:36570.
[18] van Dijken JWV, Sunnegrdh-Grnberg K. Fiber-reinforced
packable resin composites in Class II cavities. J Dent
2006;34:7639.
[19] AGY. High Strength Glass Fibers. Technical Paper; 1996.
[20] Cheng TH, Jones FR, Wang D. Effect of ber conditioning on
the interfacial shear strength of glass-ber composites.
Compos Sci Technol 1993;48:8996.
[21] Wang D, Jones FR, Denison P. Surface analytical study of the
interaction between gamma amino propyl triethoxysilane
and E-glass surface. Part I: time-of-ight secondary ion mass
spectrometry. J Mater Sci 1992;46:3648.
[22] Itota T, Carrick TE, Yoshiyama M, McCabe JF. Fluoride release
and recharge in giomer, compomer and resin composite.
Dent Mater 2004;20:78995.
[23] Guth JH. The Mechanism of Asbestos-induced
Carcinogenesis: Calcium and Plasma Membrane Integrity.

1244

[24]

[25]

[26]

[27]

[28]

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 12341244

Paper presented to the Virginia Academy of Sciences, 59th

Annual Meeting, Norfolk, VA, May 14; 1981.
Lockey JE, Roggli VL, Hilbert TJ, Rice CH, Levin LS, Borton EK,
et al. Biopersistence of refractory ceramic ber in human
lung tissue and a 20-year follow-up of radiographic pleural
change in workers. J Occup Environ Med 2012;54:7818.
Chen L, Yu Q, Wang Y, Li H. BisGMA/TEGDMA dental
composite containing high aspect-ratio hydroxyapetite
nano bers. Dent Mater 2011;27:118795.
Alander P, Lassila LVJ, Vallitu P. The span length and
cross-section design affect values of strength. Dent Mater
2005;21:34753.
Garoushi S, Vallittu PK, Lassila L. Short glass ber reinforced
restorative composite resin with semi-inter penetrating
polymer network matrix. Dent Mater 2007;23:135662.
Kardos JL. Critical issues in achieving desirable mechanical
properties for short ber composites. Pure Appl Chem
1985;57:16517.

[29] Garoushi S, Sailynoja E, Vallittu P, Lassila L. Physical

properties and depth of cure of a new short ber reinforced
composite. Dent Mater 2013;29:83541.
[30] Opdam NJM, Roeters JJM, Joosten M, Veeke O. Porosities and
voids in class I restorations placed by six operators using a
packable or syringable composite. Dent Mater 2002;18:5863.
[31] Cooke TF. Fiber reinforcement high performance. John Wiley
and Sons; 1993. p. 21732.
[32] Callister WD. Materials science and engineering: an
introduction. John Wiley & Sons; 2007. p. 585600.
[33] Zhu YT, Valdez JA, Shi N, Lovato ML, et al. Inuence of
reinforcement morphology on the mechanical properties of
short bre composites. In: Processing of Metals and
Advanced Materials: Modeling, Design and Properties. The
Minerals, Metals & Materials Society; 1998. p. 25160.
[34] Brandt AM. Fibre reinforced cement-based composites after
over 40 years of development in building and civil
engineering. Compos Struct 2008;86:39.