Elsevier Editorial System(tm) for Composites

Part A
Manuscript Draft

Manuscript Number:

Title: Vibration Damping of Flax Fibre-reinforced Polypropylene

Composites

Article Type: Research Paper

Keywords: A. Flax-polypropylene composites; B. Damping; B. Loss factor;

B. Natural frequency

Corresponding Author: Mr. Md. Zillur Rahman,

Corresponding Author's Institution: The University of Auckland

First Author: Md. Zillur Rahman

Order of Authors: Md. Zillur Rahman; Krishnan Jayaraman; Brian R Mace

Abstract: This work investigates the effects of fibre content and fibre
orientation on the damping properties of flax fibre-reinforced
polypropylene composites. Laminates of various fibre contents were
manufactured by a vacuum bagging process; their dynamic properties were
then found from vibration measurements of beam test specimens using an
impulse hammer technique to frequencies of 1 kHz. The circle-fit method
and the Newton divided difference formula were used to estimate the
natural frequencies and loss factors. The damping estimates were also
investigated using a "carpet" plot. The results showed that composites
containing 45°, 60° and 90° fibre orientation exhibit approximately the
same natural frequencies. Composites with differing fibre orientations
exhibited different loss factors, with the loss factor generally lying in
the range of 2-7%. These outcomes indicate that this composite could be a
commercially viable material for applications in which noise and
vibration are significant issues and where significant damping is
required.

Suggested Reviewers: Hao Wang PhD

Professor, University of Southern Queensland
Hao.Wang@usq.edu.au

Simon Bickerton PhD

Professor, Mechanical Engineering, The University of Auckland
s.bickerton@auckland.ac.nz

Marie-Joo Le Guen PhD

Materials scientist, Scion, Rotorua
mariejoo.leguen@scionresearch.com

Edgar Alejandro Marañon Leon PhD

Associate Professor, Mechanical Engineering, Universidad de los Andes
emaranon@uniandes.edu.co
Manuscript
Click here to view linked References

1 Vibration Damping of Flax Fibre-reinforced Polypropylene

2
3
4 Composites
5
6
7
8
Md. Zillur Rahman(*), Krishnan Jayaraman and Brian Richard Mace
9
10
11 Department of Mechanical Engineering, University of Auckland, Private Bag 92019,
12
13 Auckland 1142, New Zealand
14
15
16 (*)Email: mrah082@aucklanduni.ac.nz
17
18
19 Tel: +64 9 373 7599 ext. 89840
20
21
22
23
24
Abstract
25
26
27
28 This work investigates the effects of fibre content and fibre orientation on the damping
29
30 properties of flax fibre-reinforced polypropylene composites. Laminates of various fibre
31
32
contents were manufactured by a vacuum bagging process; their dynamic properties
33
34
35 were then found from vibration measurements of beam test specimens using an impulse
36
37 hammer technique to frequencies of 1 kHz. The circle-fit method and the Newton
38
39
40 divided difference formula were used to estimate the natural frequencies and loss
41
42 factors. The damping estimates were also investigated using a “carpet” plot. The results
43
44
45 showed that composites containing 45°, 60° and 90° fibre orientation exhibit
46
47 approximately the same natural frequencies. Composites with differing fibre
48
49
50
orientations exhibited different loss factors, with the loss factor generally lying in the
51
52 range of 2-7%. These outcomes indicate that this composite could be a commercially
53
54 viable material for applications in which noise and vibration are significant issues and
55
56
57 where significant damping is required.
58
59
60
61
62 1
63
64
65
 1 Keywords: A. Composite materials, A. Flax-polypropylene composites, B. Damping, B.
2 Loss factor, B. Natural frequency
3
4
5
6
7 1. Introduction
8
9
10
11
In areas such as the automotive industry (electric/hybrid cars), aerospace and sport,
12
13 reduction in the mass of a structure is becoming increasingly important and desirable, as
14
15 it relates directly to cost, performance and better fuel efficiency. However, reduction of
16
17
18 mass can cause a structure to be more vulnerable to audible noise and vibration. To
19
20 attenuate the vibrations to a desirable level, damping materials may be added to a
21
22
23 structure to dissipate vibrational energy. Furthermore, increased environmental
24
25 consciousness, the reduction of available non-renewable sources and the problems
26
27
28 inherent to the disposal of waste materials, have all led to a growing interest in the use
29
30 of natural fibre based composites for such purposes.
31
32
33 One particular advantage of using reinforcing agents such as natural fibres in
34
35
36 composites is their multi-scale structure. These natural fibres are composed of
37
38 elementary fibres bonded by a pectin matrix, where each of the elementary fibres can be
39
40
41
considered as a composite structure. This structure consists of stiff cellulose
42
43 microfibrills reinforced in a hemicellulose and lignin matrix [1]. These fibres are
44
45 viscoelastic and hierarchical in nature which contributes to the dissipation of energy [2,
46
47
48 3]. Furthermore, interfacial friction between fibre and matrix provides additional
49
50 damping. Therefore, the embedding of natural fibres into polymeric matrix materials
51
52
53 can have multifunctional capabilities [4] such as vibration control, energy dissipation,
54
55 and heat dissipation along with a high stiffness to weight ratio.
56
57
58
59
60
61
62 2
63
64
65
 1 To investigate the vibration damping behaviour of NFPCs (natural fibre-reinforced
2
3
4 polymer composites), experimental dynamic mechanical analysis (DMA) has been
5
6 extensively used in the past decade [2, 3, 5-13]. Duc et al. [2, 3] measured the damping
7
8 of flax, carbon and glass fibre-reinforced epoxy composites and showed that flax fibre
9
10
11 composites possess a higher damping than do glass and carbon fibre composites.
12
13 Wielage et al. [5] also found that the loss factors of flax and hemp fibre-reinforced
14
15
16 polypropylene (PP) composites are greater than those of glass fibre composites. The
17
18 loss factors increase with the twist angle of the flax yarns and the crimp in the flax
19
20
21 fabrics; this is due to increasing intra-yarn and inter-yarn friction, respectively, as stated
22
23 in [2, 6]. Idicula et al. [7] studied the effect of fibre volume fraction on the damping of
24
25
26
short banana-sisal hybrid fibre-reinforced polyester composites and found that
27
28 composites with a fibre volume fraction of 0.40 display the maximum damping. Joseph
29
30 et al. [8] and Nair et al. [9] reported that the incorporation of short sisal fibre reduces the
31
32
33 loss factors of the composites when compared with the polypropylene and polystyrene
34
35 samples, respectively. The low frequencies result in less damping of the banana fibre-
36
37
38 reinforced polyester composites compared to measurements at higher frequencies, as
39
40 described by Pothan et al. [10]. In another study, Le Guen et al. [11] reported that the
41
42
43
damping of polyol-treated flax fibre-reinforced epoxy composites is higher than that of
44
45 non-treated flax fibre composites.
46
47
48 In the aforementioned references, except for [1, 4], DMA tests were generally
49
50
51 conducted at various temperatures with a specific frequency less than or equal to 100
52
53 Hz. Etaati et al. [14] performed measurements on hemp fibre-reinforced PP composites
54
55 with excitation frequencies of up to 200 Hz. The maximum loss factor was found for
56
57
58 30% by weight hemp fibre composites, and the damping exhibited a slight dependence
59
60
61
62 3
63
64
65
 1 on frequency below 30 Hz. However, it is necessary to investigate a broader frequency
2
3
4 range to characterise the dynamic behaviour of a material, in particular the stiffness and
5
6 loss factor, as audible frequencies range from 20 to 20,000 Hz. In order to reduce noise
7
8 and vibration in transport, NFPCs can be considered as potentially useful materials for
9
10
11 interior application. In vehicles, for example, frequencies of up to a few kHz are often
12
13 important, and the frequency range from 100-1000 Hz is particularly significant in
14
15
16 causing fatigue for both driver and passengers [15]. This range also includes much of
17
18 the energy associated with internally and externally radiated noise in transport such as
19
20
21 cars and trains. Hence, the ability to add damping over a wider frequency range is
22
23 important, and this provides the motivation in this paper to measure damping over a
24
25
26
wider frequency range of up to 1000 Hz.
27
28
29 Some studies [12, 13, 16] have estimated the natural frequencies and loss factors of
30
31 NFPCs. Landro and Lorenzo [12, 13] examined the dynamic behaviour of natural fibre
32
33
34
mat (flax, kenaf and hemp fibres were used in the mat, 50% by weight) reinforced PP
35
36 composites over excitation frequencies of up to 4 kHz. An increase in damping with
37
38 frequency was seen. Kumar et al. [16] reported only the first three natural frequencies
39
40
41 and the loss factors associated with short sisal and banana fibre-reinforced polyester
42
43 composites for various fibre lengths (3 mm, 4 mm and 5 mm) and contents (30%, 40%
44
45
46 and 50% by weight). They demonstrated that the natural frequencies of the composites
47
48 are not affected significantly by variations in fibre length. For a constant fibre length (3
49
50
51 mm), two kinds of damping trends were observed with the increase in banana and sisal
52
53 fibre contents in the case of the first mode: for the former, damping decreases; for the
54
55 latter, it increases. In those studies, the estimation of loss factor was carried out using
56
57
58 the half-power bandwidth method, also known as the peak-picking method. This
59
60
61
62 4
63
64
65
 1 method produces an overestimation of damping; this is due to the assumption that only
2
3
4 a single mode contributes to the response around each resonance frequency. However,
5
6 off-resonant modes can contribute a significant amount to the total response at any
7
8 resonance. The damping estimation errors can be up to 20% or more for a multiple-
9
10
11 degree-of-freedom system [17] depending on the mode number, modal overlap and loss
12
13 factor. On the other hand, the single-degree-of-freedom (SDOF) circle-fit method [18]
14
15
16 significantly improves the accuracy of damping estimation compared to the half-power
17
18 bandwidth method, as it considers the frequency range in the close vicinity of a natural
19
20
21 frequency. This approach is particularly accurate when the modal overlap is less than
22
23 one. In this approach, the total contribution of all the off-resonant modes is assumed to
24
25
26
vary very slowly over the narrow frequency band around a specific resonance. It also
27
28 provides more robust estimates of loss factor because a number of estimates can be
29
30 found from responses at different frequencies around each resonance and are then
31
32
33 averaged. In addition, for both techniques, frequency response is typically measured
34
35 using a digital spectral analysis with a given frequency resolution which may also lead
36
37
38 to inaccuracies. Thus, in this paper, natural frequencies and loss factors are estimated
39
40 using the SDOF circle-fit method [18] and the Newton divided difference formula [19,
41
42
43
20].
44
45
46 One further problem associated with experimental measurements reported in the
47
48 literature [12-14, 16, 21, 22] is that in those studies accelerometers were attached to the
49
50
51 specimens. However, the accelerometer and its lead add mass and damping to the
52
53 structure and affect the estimates of natural frequencies and loss factors. Therefore, in
54
55 this work, a non-contacting transducer such as a laser vibrometer is used.
56
57
58
59
60
61
62 5
63
64
65
 1 To date, a number of researches have been carried out on the vibration damping of
2
3
4 NFPCs, but few studies have been conducted to ascertain the structural damping levels
5
6 of NFPCs over a broader frequency range. The present work is therefore an attempt to
7
8 measure the dynamic characteristics of NFPCs over a broader frequency range. The
9
10
11 dynamic characterisation concerns various composite beams of different lengths with
12
13 three different volume fractions and five different orientations of flax fibre. The quality
14
15
16 of the damping estimates from the “carpet” plots is also investigated.
17
18
19 The experimental details and data processing are described in section 2. In section 3,
20
21 experimental results involving measurements taken from various composite beams are
22
23
24 presented and discussed; an assessment of the quality of the estimated damping is also
25
26 given. Section 4 contains the concluding observations for the study.
27
28
29 2. Experimental details
30
31
32
33
2.1. Materials
34
35
36 Sheets of PP random copolymer (MOPLEN RP241G), with a melt flow rate (MFR) of
37
38 1.5 g/10 min determined by ISO 1133 and a thickness of 0.38 mm were used as matrix
39
40
41
material. The polypropylene sheets were produced by Lyondell Basell Industries and
42
43 supplied by Field International Ltd., Auckland, New Zealand. The mechanical
44
45 properties of PP are listed in Table 1. Unidirectional flax fabric (FlaxPly UD180) with a
46
47
48 nominal specific weight of 180 g/m2 and density of 1410 kg/m3 was used as
49
50 reinforcement. The flax fabric (42.5 yarns/cm (warp) and 3 yarns/cm (weft)) was
51
52
53 supplied by Lineo, Meulebeke, Belgium. The weight distribution of flax fabric in the
54
55 warp and weft directions was 95.5% and 4.5%, respectively. A micrograph of the flax
56
57
58
fabric is also shown in Fig. 1.
59
60
61
62 6
63
64
65
 1 2.2. Manufacturing of composites
2
3
4 Flax fabrics were dried for 24 hours at 70°C in a vacuum dryer (Squaroid duo-vac
5
6
7 vacuum oven) to reduce the moisture content. The vacuum bagging technique (Fig. 2)
8
9 was used to manufacture the composite samples. This technique uses atmospheric
10
11
12 pressure for holding the laminate in place during the cure cycle. The consolidation
13
14 temperature of 190°C was chosen based on the differential scanning calorimetry curve
15
16
of PP, as shown in Fig. 3, which was greater than the second peak melting temperature
17
18
19 of 149°C. Dry flax fabrics and PP sheets were interleaved by a hand lay-up process and
20
21 placed on an aluminium plate. Then, a peel ply was used to separate the breather from
22
23
24 the laminate, and the breather was employed to ensure all the air inside the vacuum bag
25
26 could be drawn into a vacuum port. After sealing the materials stack, all the air was
27
28
29 evacuated from inside the vacuum bag using a vacuum pump. The mould was
30
31 subsequently placed inside the Elecfurn (FAC 100) oven and heated to a temperature of
32
33
34
190°C for 1 hour. After this, the mould was cooled to a room temperature of 25°C. The
35
36 temperature-pressure cycle (cycle 1) was applied according to reference [23]. The size
37
38 of the panels was nominally 500 mm x 600 mm, with a target thickness of 3 mm and
39
40
41 target fibre volume fractions of 0.31, 0.40 and 0.50. Corresponding to each fibre
42
43 composition, beams (450 mm x 20 mm) with fibre orientations of 0°, 30°, 45°, 60° and
44
45
46 90° were cut from the panels using an automatic saw. Neat PP samples were also
47
48 manufactured for comparison.
49
50
51 2.3. Impulse hammer technique
52
53
54 An impulse hammer technique was used to measure the frequency response of the
55
56
57 NFPC samples at an ambient temperature of ~21 °C. One end of the composite beam
58
59
60
61
62 7
63
64
65
 1 was clamped into a fixed support and the other end was free to vibrate, as illustrated in
2
3
4 Fig. 4. The excitation was provided by an impact hammer (PCB model: 086E80) with a
5
6 soft tip at a distance of 5 mm from the free end of the beam, and the beam response was
7
8 measured by a laser vibrometer (Polytec model: PDV-100) at a distance of 20 mm from
9
10
11 the free end of the beam. The excitation and response signals were then processed by a
12
13 spectrum analyser which gives the frequency response functions. Subsequent data
14
15
16 processing was performed in MATLAB [24]. The estimation of natural frequencies and
17
18 loss factors was made on composite beams of lengths 300, 350, and 400 mm with
19
20
21 different fibre volume fractions and orientations. The results are reported for beams of
22
23 differing lengths to provide a range of resonance frequencies.
24
25
26 2.4. Frequency response functions
27
28
29 The accelerance (acceleration per unit force) was measured over a frequency range to 1
30
31
32 kHz. Fig. 5, Fig. 6 and Fig. 7 show examples of the magnitude of the measured
33
34 accelerance for flax/PP samples of various fibre volume fractions and orientations, and
35
36
37 the neat PP sample. There are a number of resonance peaks for the various modes of
38
39 vibration in the frequency range, with the natural frequencies being determined by the
40
41
42
stiffness and length of the beam. The data in the vicinity of these peaks was then post-
43
44 processed to determine the natural frequencies and the loss factors.
45
46
47
48
49
50 2.5. Modal analysis using the SDOF circle-fit method
51
52
53 The extraction of modal parameters such as natural frequencies and loss factors from
54
55
56 measured frequency responses (see Fig. 5, Fig. 6 and Fig. 7) was performed using the
57
58 SDOF circle-fit method [18]. An illustrative example is shown in Fig. 8. A portion of
59
60
61
62 8
63
64
65
 1 the data in a narrow frequency range around each resonance was analysed. A circle was
2
3
4 fitted through the use of a least-squares error fit to the data when plotted in the complex
5
6 plane; the centre and the radius of the circle were estimated. The Newton divided
7
difference formula was used to determine the location of the natural frequency (r ) and
8
9
10
11 its value corresponding to the maximum rate of change of phase [19, 20]. The loss
12
13
14 factor ( r ) was estimated from the frequency response measurements at frequencies a
15
16
17 and b above and below the natural frequency using the expression [18].
18
19
20 a2  b2
21 r  (1)
22  2  tan   a   tan  b  
  2   2  
  
r
23
24
25
26
where  a ,b are the angles between the radii from the centre of the circle to the natural
27
28
29 frequency and the frequency response at the chosen frequencies a and b ,
30
31
32 respectively. The mean loss factors were then calculated from 100 estimates by
33
34
35
considering 20 data points, 10 data points below the natural frequency and 10 data
36
37 points above the natural frequency.
38
39
40 3. Results and discussion
41
42
43 3.1. Natural frequencies
44
45
46 The variations of natural frequency with fibre orientation for the different fibre volume
47
48
49 fractions are shown in Fig. 9, Fig. 10 and Fig. 11. In general, the natural frequencies of
50
51 the neat PP beam are lower than those of the composite beams, and this is expected as
52
53
54 the stiffness of the neat PP sample is lower. In regard to the fibre orientations of 0° and
55
56 90° samples, the natural frequencies increase with higher fibre content. This is due to
57
58
59
the presence of a higher proportion of fibres producing greater stiffness, and hence the
60
61
62 9
63
64
65
 1 natural frequency is increased. For the composite orientations of 30°, 45° and 60°, the
2
3
4 natural frequency either decreases by a small amount or is almost constant with
5
6 increasing fibre content.
7
8
9 This is consistent with the results obtained from static three-point bend tests (see Fig.
10
11
12 12); the flexural moduli decreased somewhat for these fibre orientations with increasing
13
14 fibre content. The flexural moduli were measured in accordance with the standard
15
16
ASTM D760-10 using an Instron 5567 with a 10 kN load cell. The average value and
17
18
19 standard error obtained from at least five specimens (nominal size of each specimen: 75
20
21 mm x 12.7 mm x thickness) are given.
22
23
24
As the angle between the longitudinal and fibre axes increases, the drop in natural
25
26
27 frequencies is substantial. However, the rate of decrease in natural frequencies reduces
28
29 when the angle between them is greater than 30°. Increasing the angle of fibres from 0°
30
31
32 to 60° reduces the natural frequency by 63.72% (from 21.64 Hz to 7.85 Hz) in the case
33
34 of first mode of a fibre volume fraction of 0.31. Incorporation of fibres (V f  0.50) into
35
36
37 PP increases the natural frequency of the 0° oriented samples from approximately 7.59
38
39
40
Hz to approximately 26.95 Hz. The highest natural frequency occurs for 0° fibre
41
42 orientation because the fibres are stiffer in tension than the matrix. Kumar et al. [16]
43
44 also reported that the more rigid specimens have higher natural frequencies. The lowest
45
46
47 natural frequency occurs for 90° fibre orientation in the instance of two distinct fibre
48
49 volume fractions of 0.31 and 0.40; in contrast, in the case of 0.50 fibre volume fraction,
50
51
52 the maximum natural frequency occurs at 0° fibre orientation and the minimum at 60°,
53
54 although the variation between 60°and 90° is negligible.
55
56
57
58
59
60
61
62 10
63
64
65
 1 The magnitude of the peak value of the response at resonance decreases with increasing
2
3
4 beam length, as illustrated in Fig. 13. The modal density increases with increasing
5
6 length. A bending-twisting mode (see Fig. 7) appears for beam lengths of 300 mm, 350
7
8 mm, 400 mm, and a fibre volume fraction of 0.50 and orientation of 30°. It originates
9
10
11 from the coupling between bending and twisting when fibre orientation is different from
12
13 0° or 90°. Moreover, the coupling effect reduces the stiffness resulting in lower natural
14
15
16 frequencies [25].
17
18
19 3.2. Loss factors
20
21
22 3.2.1 Effect of fibre content
23
24
25 Fig. 14 shows the variation of mean loss factor with frequency of flax/PP composite
26
27
28 beams for a fibre volume fraction of 0.31, and neat PP beams. Variations for fibre
29
30 volume fractions of 0.40 and 0.50 are shown in Fig. 15 and Fig. 16, respectively. Linear
31
32
33
fits are also shown for each set of data.
34
35
36 The loss factor of neat PP is approximately 0.06 at low frequencies. This value matches
37
38 with the value reported by a number of authors [26, 27]. As the frequency increases, the
39
40
41
loss factor of neat PP increases, rising to 0.090 at approximately 500 Hz.
42
43
44 With increased fibre content of 40% and 50% by volume, flax/PP samples exhibit
45
46 approximately the same loss factors (loss factors lie predominantly in the range of 4-6%
47
48
49
for the fibre orientations of 30°, 45°, 60° and 90°, and 2-2.5% for the fibre orientation of
50
51 0°) as 31% by volume, despite a decrease in the amount of the viscoelastic polymer. In
52
53 general, as fibre content increases, the loss factor decreases. However, no significant
54
55
56 difference in the loss factors was observed with the addition of an increase in the
57
58 amount of flax fibre. This is probably due to the existence of more fibres offering more
59
60
61
62 11
63
64
65
 1 fibre-matrix interfaces. This results in an increase in the number of energy dissipation
2
3
4 sites. This is also supported by the statement of Chauhan et al. [28] that increasing filler
5
6 content can contribute to the creation of large number of interfacial areas which
7
8 increases vibrational energy dissipation. A similar observation was reported by Liang
9
10
11 and Tjong [29]. Moreover, this is due to the internal structure of the fibre which induces
12
13 high internal friction [21]; in particular, this occurs between the cellulose and
14
15
16 hemicellulose in each wall and the friction between the cell walls [2, 3, 6]. The addition
17
18 of natural fibres also contributes to energy dissipation, even at room temperature, and
19
20
21 results in the significant damping of the composites over a wide range of frequency
22
23 [12]. In addition, high damping can be attributed to the presence of voids (e.g. the
24
25
26
intrinsic porosity of natural fibres) and the viscoelastic nature of matrix and
27
28 reinforcement.
29
30
31 The loss factor of composites is, however, lower than that of neat PP because of the
32
33
34
lower amount of viscoelastic PP. This coincides with the trend reported by Chandra et
35
36 al. [30] that high content of resin leads to higher damping due its viscoelastic nature.
37
38 The stiffness is strongly dependent on the presence of fibre, while the loss factor is
39
40
41 dominated by the matrix; this results in decreased loss factors due to the incorporation
42
43 of flax fibres into PP. A similar observation has also been reported by other authors [3,
44
45
46 14, 28, 30-32]. However, interfacial damping ameliorates this. When compared to the
47
48 base PP samples, at fibre volume fraction of 0.31, the flexural stiffness of the composite
49
50
51 material increases by about 270% (from 1.07 GPa to 3.96 GPa, as shown in Fig. 12),
52
53 whereas the loss factor decreases by about 30% (from 0.078 to 0.054) in the case of 30°
54
55 fibre oriented samples.
56
57
58
59 3.2.2 Effect of fibre orientation
60
61
62 12
63
64
65
 1 Fig. 17 depicts the variation of mean loss factor with fibre orientation of flax/PP
2
3
4 composite beams for a fibre volume fraction of 0.31, and neat PP beams. Variations for
5
6 fibre volume fractions of 0.40 and 0.50 are shown in Fig. 18 and Fig. 19, respectively.
7
8
9 The effect of reinforcement orientation on the damping of the composites is noticeable.
10
11
12 The 0° fibre oriented samples demonstrate sharper resonance peaks (see Fig. 5, Fig. 6
13
14 and Fig. 7), while 30°, 45°, 60° and 90° fibre oriented samples exhibit broader peaks
15
16
with the exception of the first two peaks. This indicates that the damping is greater for
17
18
19 such orientations.
20
21
22 The loss factor increases with increasing fibre orientation up to 45°, and it can then be
23
24
seen to decrease slightly as the fibre orientation increases to above 45°. Maximum
25
26
27 damping was obtained for the fibre orientation of 45° in the case of all fibre volume
28
29 fractions, and the damping was found to lie in the range of 4-7% for this fibre
30
31
32 orientation. A similar trend was observed by Ying et al. [33] and Adams and Maheri
33
34 [34] in the cases of carbon fibre-reinforced epoxy, and carbon and glass fibre-reinforced
35
36
37 plastic composites, respectively. This is because the total energy is dominated by the in-
38
39 plane shear strain energy [35]; the in-plane shear strain energy is maximised at this fibre
40
41
42
orientation in NFPCs. The maximum loss factor of 0.071 was observed for a 300 mm
43
44 beam (V f  0.31 and   45) at approximately 500 Hz. The average loss factor for the
45
46
47 45° oriented samples showed a significant increase in loss factor of around 150% (from
48
49 0.021 to 0.052) when compared to the 0° fibre oriented samples. At very low frequency,
50
51
52 fibre orientations other than 0° fibre orientation exhibited almost the same damping.
53
54 The damping is minimal and the stiffness is at its maximum at 0°, i.e., in the fibre
55
56
57 direction. This is reasonable, since the fibres play a dominant role at 0° fibre orientation
58
59 and fibres exhibit less damping and a much higher stiffness than the matrix material. As
60
61
62 13
63
64
65
 1 the orientation angle increases, the general trend is for the damping to increase (up to a
2
3
4 certain angle) and the stiffness to decrease. The bending-twisting modes exhibit a loss
5
6 factor of 0.032 on average; this is less than other modes of the same sample.
7
8
9 3.2.3 Effect of frequency
10
11
12 Damping increases when frequency increases for a given fibre volume fraction. The
13
14
15 increase is in the range of 25-60% over the frequency range to 1 kHz for all fibre
16
17 volume fractions and orientations. The increase in energy dissipation with frequency
18
19
20 may come from fibres and/or fibre/matrix interactions.
21
22
23 The fibre orientation (30°, 45°, 60° and 90°) of the sample affects the loss factor more
24
25 than the fibre content (31%, 40% and 50% by volume). Overall, the loss factor of
26
27
28 NFPCs is in the range of 2-7% for all studied fibre volume fractions and orientations. It
29
30 can be inferred from the above discussion that natural fibres possess a significant
31
32
33
vibration damping capacity.
34
35
36 3.3. Diagnostic of the quality of the damping estimates
37
38
39 The loss factor estimated using Eq. (1) depends on the chosen frequencies a and b in
40
41
42 addition to the natural frequency (r ) . Different choices of a and b give different
43
44 estimates of loss factor. The variation in the individual estimates of loss factor for many
45
46
47 combinations of the selected frequencies can be seen from the damping “carpet” plots,
48
49 with two examples shown in Fig. 20 and Fig. 21. The plots are presented for the first
50
51
52 and second modes of a beam (beam length = 300 mm, V f  0.50 and   0) using all
53
54
55 combinations of the 20 selected frequencies below and above the natural frequency.
56
57 This gives rise to 100 estimates of loss factor from which are obtained the mean and
58
59
60
61
62 14
63
64
65
 1 variance of the estimates. For a perfect case, the surface ought to be smooth, flat and
2
3
4 level. The surfaces of the plots are almost horizontal, smooth and flat, apart from some
5
6 tilt and unevenness. The coefficients of variation of the 100 estimates are 2.27% and
7
8 3.58% for mode 1 and 2, respectively. These are small and indicate that the estimates
9
10
11 are of good quality. On the whole, the coefficients of variation are in the range of 1.04-
12
13 9.92%, 1.40-10.53% and 1.17-11.11% in the case of fibre content of 31%, 40% and
14
15
16 50% by volume, respectively. Relatively large coefficients of variation are observed in
17
18 the case of the highest modes. In relation to this deviation, some possible measurement
19
20
21 errors include measurement noise, clamping pressure, air damping and non-uniformity
22
23 in the laminate (voids, variations in thickness and improper bonding).
24
25
26
27
28
29 4. Concluding remarks
30
31
32
33
The effect of the fibre content, fibre orientation, and frequency on damping in NFPCs
34
35 was estimated from vibration measurements. The SDOF circle-fit method and the
36
37 Newton divided difference formula provide accurate estimates of natural frequencies
38
39
40 and loss factors. Changes in fibre content and orientation produce different natural
41
42 frequencies and loss factors for the same geometry and boundary condition. This offers
43
44
45 an additional freedom for the designing of a composite laminate – allowing for the
46
47 prospect of alternating the fibre orientation to affect the structure’s stiffness and
48
49
50
damping. This knowledge also makes the use of these materials attractive as it is
51
52 feasible to increase damping without increasing mass or changing geometry.
53
54
55 Of all the parameters, fibre orientation has the most significant impact on damping. At
56
57
58
each fibre volume fraction, the loss factor increases up to 4-7% for fibre orientation up
59
60
61
62 15
63
64
65
 1 to 45° and it then decreases slightly. The loss factor generally lies in the range of 2-7%,
2
3
4 irrespective of various fibre volume fraction, fibre orientation and frequency. The
5
6 damping “carpet” plots indicate the quality of the estimates of loss factor. The
7
8 coefficient of variation of the loss factor estimated at each resonance was in the range of
9
10
11 1.04-11.11%.
12
13
14 Given this, it may be inferred that the high damping capability of NFPCs with low
15
16
weight make these materials attractive for applications in cases where high damping is
17
18
19 sought.
20
21
22
23
24
25 Acknowledgements
26
27
28 The first author acknowledges the financial support provided by the Department of
29
30
31 Mechanical Engineering, The University of Auckland.
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 16
63
64
65
 1 References
2
3 [1] Li Y, Luo Y, Han S. Multi-scale structures of natural fibers and their applications in
4 making automobile parts. Journal of Biobased Materials and Bioenergy. 2010;4(2):164-
5
6
71.
7 [2] Duc F, Bourban PE, Manson JAE. The role of twist and crimp on the vibration
8 behaviour of flax fibre composites. Composites Science and Technology. 2014;102:94-
9 9.
10 [3] Duc F, Bourban PE, Plummer CJG, Månson JAE. Damping of thermoset and
11
12 thermoplastic flax fibre composites. Composites Part A: Applied Science and
13 Manufacturing. 2014;64:115-23.
14 [4] Gibson RF. A review of recent research on mechanics of multifunctional composite
15 materials and structures. Composite Structures. 2010;92(12):2793-810.
16
[5] Wielage B, Lampke T, Utschick H, Soergel F. Processing of natural-fibre reinforced
17
18 polymers and the resulting dynamic–mechanical properties. Journal of Materials
19 Processing Technology. 2003;139(1–3):140-6.
20 [6] Duc F, Bourban PE, Manson JAE. Dynamic mechanical properties of epoxy/flax
21 fibre composites. Journal of Reinforced Plastics and Composites. 2014;33(17):1625-33.
22
23
[7] Idicula M, Malhotra SK, Joseph K, Thomas S. Dynamic mechanical analysis of
24 randomly oriented intimately mixed short banana/sisal hybrid fibre reinforced polyester
25 composites. Composites Science and Technology. 2005;65(7–8):1077-87.
26 [8] Joseph P, Mathew G, Joseph K, Groeninckx G, Thomas S. Dynamic mechanical
27 properties of short sisal fibre reinforced polypropylene composites. Composites Part A:
28
29 Applied Science and Manufacturing. 2003;34(3):275-90.
30 [9] Manikandan Nair K, Thomas S, Groeninckx G. Thermal and dynamic mechanical
31 analysis of polystyrene composites reinforced with short sisal fibres. Composites
32 Science and Technology. 2001;61(16):2519-29.
33
34
[10] Pothan LA, Oommen Z, Thomas S. Dynamic mechanical analysis of banana fiber
35 reinforced polyester composites. Composites Science and Technology. 2003;63(2):283-
36 93.
37 [11] Le Guen M-J, Newman RH, Fernyhough A, Staiger MP. Tailoring the vibration
38 damping behaviour of flax fibre-reinforced epoxy composite laminates via polyol
39
40 additions. Composites Part A: Applied Science and Manufacturing. 2014;67:37-43.
41 [12] Di Landro L, Lorenzi W. Static and dynamic properties of thermoplastic
42 matrix/natural fiber composites. Journal of Biobased Materials and Bioenergy.
43 2009;3(3):238-44.
44
45
[13] Di Landro L, Lorenzi W. Mechanical properties and dynamic mechanical analysis
46 of thermoplastic‐natural fiber/glass reinforced composites. Macromolecular Symposia:
47 Wiley Online Library; 2009. p. 145-55.
48 [14] Etaati A, Mehdizadeh SA, Wang H, Pather S. Vibration damping characteristics of
49 short hemp fibre thermoplastic composites. Journal of Reinforced Plastics and
50
51 Composites. 2013;33(4):330-41.
52 [15] Dias T, Monaragala R. Sound absorbtion in knitted structures for interior noise
53 reduction in automobiles. Measurement Science and Technology. 2006;17(9):2499.
54 [16] Senthil Kumar K, Siva I, Jeyaraj P, Winowlin Jappes JT, Amico SC, Rajini N.
55
Synergy of fiber length and content on free vibration and damping behavior of natural
56
57 fiber reinforced polyester composite beams. Materials & Design. 2014;56:379-86.
58
59
60
61
62 17
63
64
65
 1 [17] Wang J-T, Jin F, Zhang C-H. Estimation error of the half-power bandwidth method
2 in identifying damping for multi-DOF systems. Soil Dynamics and Earthquake
3
4 Engineering. 2012;39:138-42.
5 [18] Ewins DJ. Modal Testing: Theory, Practice and Application. Baldock,
6 Hertfordshire, England: Research Studies Press Ltd.; 2000.
7 [19] Maia NMM, E Silva JMM. Theoretical and experimental modal analysis. Baldock,
8 Hertfordshire, England: Research Studies Press Ltd.; 1997.
9
10 [20] E Silva JM, Maia N. Single mode identification techniques for use with small
11 microcomputers. Journal of Sound and Vibration. 1988;124(1):13-26.
12 [21] Assarar M, Zouari W, Sabhi H, Ayad R, Berthelot J-M. Evaluation of the damping
13 of hybrid carbon–flax reinforced composites. Composite Structures. 2015;132:148-54.
14
15
[22] Cheour K, Assarar M, Scida D, Ayad R, Gong X-L. Effect of water ageing on the
16 mechanical and damping properties of flax-fibre reinforced composite materials.
17 Composite Structures. 2016;152:259-66.
18 [23] Kazmi S, Das R, Jayaraman K. Sheet forming of flax reinforced polypropylene
19 composites using vacuum assisted oven consolidation (VAOC). Journal of Materials
20
21 Processing Technology. 2014;214(11):2375-86.
22 [24] MATLAB R2013b. 8.2.0.701 ed. Natick, Massachusetts: The MathWorks Inc.
23 [25] Thornton E, Clary R. A correlation study of finite-element modeling for vibrations
24 of composite material panels. ASTM Special Technical Publication. 1974(546):111-29.
25
26
[26] Ioana C. Finegan, Gary. G. Tibbetts, Ronald F. Gibson. Modeling and
27 characterization of damping carbon nanofiber/polypropylene composites. Composites
28 Science and Technology. 2003;63:1629-35.
29 [27] Subramanian C, Deshpande SB, Senthilvelan S. Effect of reinforced fiber length on
30 the damping performance of thermoplastic composites. Advanced Composite Materials.
31
32 2011;20(4):319-35.
33 [28] Chauhan S, Karmarkar A, Aggarwal P. Damping behavior of wood filled
34 polypropylene composites. Journal of Applied Polymer Science. 2009;114(4):2421-6.
35 [29] Liang J, Li R, Tjong S. Dynamic mechanical analysis of low-density polyethylene
36
with glass beads. Journal of Thermoplastic Composite Materials. 2000;13(1):12-20.
37
38 [30] Chandra R, Singh SP, Gupta K. Damping studies in fiber-reinforced composites – a
39 review. Composites Structures. 1999;46:41-51.
40 [31] Joseph K, Thomas S, Pavithran C. Dynamic mechanical properties of short sisal
41 fiber reinforced low density polyethylene composites. Journal of Reinforced Plastics
42
43
and Composites. 1993;12(2):139-55.
44 [32] Landel RF, Nielsen LE. Mechanical properties of polymers and composites. New
45 York, USA: Marcel Dekker, Inc. ; 1993.
46 [33] Gao Y, Li Y, Hong Y, Zhang H, He X. Modeling of the damping properties of
47 unidirectional carbon fibre composites. Polymers & Polymer Composites.
48
49 2011;19(2/3):119.
50 [34] Adams R, Maheri M. Damping in advanced polymer–matrix composites. Journal
51 of Alloys and Compounds. 2003;355(1):126-30.
52 [35] Hwang SJ, Gibson RF. The use of strain energy-based finite element techniques in
53
54
the analysis of various aspects of damping of composite materials and structures.
55 Journal of Composite Materials. 1992;26(17):2585-605.
56
57
58
59
60
61
62 18
63
64
65
 1 Table 1 Mechanical properties of polypropylene.
2
3 Property Value
4
5 Density 900 kg/m3
6 Tensile modulus 1.10 GPa
7
8
Yield strength 28 MPa
9 Flexural modulus 1.05 GPa
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31 Fig. 1. Flax fabric.
32
33 To vacuum pump
34
35
36 Vacuum bag
37
38 Breather fabric
39 Peel ply
40 Laminate
41 Sealant tape
42 Aluminium plate
43
44 Fig. 2. Vacuum bagging technique.
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 19
63
64
65
 1
2 0
3
4 -0.2
5
6
-0.4
Heat flow (W/g)

7
8
9 -0.6
10
11
12 -0.8
13
14 -1
15
16
17 -1.2
0 20 40 60 80 100 120 140 160 180 200
18 Temperature (C)
19
20 Fig. 3. A differential scanning calorimetry curve of neat polypropylene.
21
22
23
24
25
26
27
28 Laser vibrometer
29
30
31
32 Impulse hammer
33
34
35
36 Composite specimen
37
38
39
40
41
42
43
44
45 Dynamic signal analyser
46
47 Fig. 4. Schematic diagram of experimental setup for an impulse hammer technique.
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 20
63
64
65
 1
4
2 10
3 Accelerance magnitude [A/F], (m/s2)/N
4
5
6
3
7 10
8
9
10
11
12 2
10
13 Flax/PP (300x20x3.07 mm3), V =0.31 ( = 0)
f

14 Flax/PP (300x20x3.07 mm3), V =0.31 ( = 30)

f
15 3 
Flax/PP (300x20x3.02 mm ), Vf=0.31 ( = 45 )
16 3 
Flax/PP (300x20x3.04 mm ), Vf=0.31 ( = 60 )
17 10
1 3 
Flax/PP (300x20x3.02 mm ), Vf=0.31 ( = 90 )
18 Neat PP (300x20x3.02 mm )
3

19
20
0 100 200 300 400 500 600 700 800 900 1000
Frequency [Hz]
21
22
23 Fig. 5. Accelerance magnitudes for flax/PP beams of different fibre orientations and a
24 fibre volume fraction of 0.31, and a neat PP beam.
25
26 4
27 10
28
Accelerance magnitude [A/F], (m/s2)/N

29
30
31 3
32 10
33
34
35
36 2
37 10
38 Flax/PP (300x20x3.00 mm3), Vf=0.40 ( = 0)

39 Flax/PP (300x20x2.94 mm3), Vf=0.40 ( = 30)

40 Flax/PP (300x20x2.90 mm3), Vf=0.40 ( = 45)
41
1 Flax/PP (300x20x3.00 mm3), Vf=0.40 ( = 60)
42 10
43 Flax/PP (300x20x3.16 mm3), Vf=0.40 ( = 90)

44 0 100 200 300 400 500 600 700 800 900 1000
45 Frequency [Hz]
46
47
48
Fig. 6. Accelerance magnitudes for flax/PP beams of different fibre orientations and a
49 fibre volume fraction of 0.40.
50
51
52
53
54
55
56
57
58
59
60
61
62 21
63
64
65
 1
4
2 10
3 Accelerance magnitude [A/F], (m/s2)/N
4
5
6
3
7 10
8
9
10
11
12 2
10
13
Flax/PP (300x20x3.22 mm3), Vf=0.50 ( = 0)
14
Flax/PP (300x20x3.39 mm3), Vf=0.50 ( = 30)
15
16 Bending-twisting coupled mode Flax/PP (300x20x3.15 mm3), Vf=0.50 ( = 45)

17 1 Flax/PP (300x20x3.28 mm3), Vf=0.50 ( = 60)

10
18 Flax/PP (300x20x3.36 mm3), Vf=0.50 ( = 90)
19
20
0 100 200 300 400 500 600 700 800 900 1000
Frequency [Hz]
21
22
23
24 Fig. 7. Accelerance magnitudes for flax/PP beams of different fibre orientations and a
25 fibre volume fraction of 0.50.
26
27
28
29
30 1000
31
0
32
33 -1000
34
35 -2000
36 168.4375Hz
87 
Imag

168.2812Hz
-3000 80 
37 168.125Hz
73 
-91  165.4688Hz
167.9688Hz -84  165.625Hz
38 65 
167.8125Hz 56 
-78  165.7813Hz
-4000 -71  165.9375Hz
39 167.6563Hz 47  -62  166.0938Hz

40 167.5Hz 37 26  -52  166.25Hz

-43 
-5000 167.3438Hz 14  -32 
 -21 
2 -10
166.4063Hz
41 167.1875Hz
167.0313Hz 166.875Hz
166.5625Hz
166.7188Hz
42 -6000 167.0009Hz
43
44 -7000
45 -6000 -4000 -2000 0 2000 4000
46 Real
47
Fig. 8. Modal circle for extracting natural frequency and loss factor:* natural frequency
48
49 (indicated by an arrow) and o discrete frequency data, angles are in degrees.
50
51
52
53
54
55
56
57
58
59
60
61
62 22
63
64
65
 1
2 1000
Flax/PP (Vf = 0.31) and Neat PP, L = 300 mm
3 900 Flax/PP (Vf = 0.31) and Neat PP, L = 350 mm
4 Flax/PP (Vf = 0.31) and Neat PP, L = 400 mm
5 800
6 700
Frequency (Hz)

7
8 600
9 500
10
11 400
12 300
13
14 200
15 100
16
17 0
0° 15° 30° 45° 60° 75° 90° PP
18 Fibre orientation
19
20
21 Fig. 9. Variation of natural frequency with fibre orientation for flax/PP beams of a fibre
22
23
volume fraction of 0.31, and neat PP beams.
24
25 1000
Flax/PP (Vf = 0.40) and Neat PP, L = 300 mm
26 Flax/PP (Vf = 0.40) and Neat PP, L = 350 mm
900
27 Flax/PP (Vf = 0.40) and Neat PP, L = 400 mm
28 800
29
700
30
Frequency (Hz)

31 600
32
500
33
34 400
35
300
36
37 200
38
100
39
40 0
0° 15° 30° 45° 60° 75° 90° PP
41 Fibre orientation
42
43
44
45 Fig. 10. Variation of natural frequency with fibre orientation for flax/PP beams of a
46 fibre volume fraction of 0.40, and neat PP beams.
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 23
63
64
65
 1
2 1000
Flax/PP (Vf = 0.50) and Neat PP, L = 300 mm
3 900 Flax/PP (Vf = 0.50) and Neat PP, L = 350 mm
4 Flax/PP (Vf = 0.50) and Neat PP, L = 400 mm
5 800 Coupling mode, L = 300 mm*
6 Coupling mode, L = 350 mm*
700
Coupling mode, L = 400 mm*
7
Frequency (Hz)

8 600
9 500
10
11 400
12 300
13
14 200
15 100
16
17 0
0° 15° 30° 45° 60° 75° 90° PP
18 Fibre orientation
19
20
21 Fig. 11. Variation of natural frequency with fibre orientation for flax/PP beams of a
22
23
fibre volume fraction of 0.50, and neat PP beams. *Natural frequencies of bending-
24 twisting coupled modes.
25
26 16
Flax/PP (Vf = 0.31)
27
Flax/PP (Vf = 0.40)
28 14 Flax/PP (Vf = 0.50)
29 Neat PP
30 12
Flexural modulus (GPa)

31
32 10
33
34 8
35
6
36
37 4
38
39 2
40
41 0
42 0° 15° 30° 45° 60° 75° 90° PP
Fibre orientation
43
44
45
46 Fig. 12. Flexural moduli for neat PP and flax/PP samples of different fibre volume
47 fractions and orientations using static three-point bending measurements.
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 24
63
64
65
 1 4
2 10
3
Accelerance magnitude [A/F], (m/s2)/N
4
5
3
6 10
7
8
9
10 2
10
11
12
13
Flax/PP(300x20x3.22 mm3), Vf=0.50 ( = 0)
14
1
15 10 Flax/PP (350x20x3.22 mm3), Vf=0.50 ( = 0)

16 Flax/PP (400x20x3.22 mm3), Vf=0.50 ( = 0)

17 0 100 200 300 400 500 600 700 800 900 1000
18 Frequency [Hz]
19
20
21 Fig. 13. Accelerance magnitudes for different flax/PP beam lengths of a fibre volume
22
23 fraction of 0.50.
24
25
26
0.1
27 Flax/PP, Vf = 0.31 ( = 0)
28 0.09 Flax/PP, Vf = 0.31 ( = 30)
29 Flax/PP, Vf = 0.31 ( = 45)
0.08
30 Flax/PP, Vf = 0.31 ( = 60)
31 0.07
Flax/PP, Vf = 0.31 ( = 90)
32
Loss factor

0.06 Neat PP
33
34 0.05
35 0.04
36
37 0.03
38 0.02
39
40 0.01
41 0
42 0 100 200 300 400 500 600 700 800 900 1000
Frequency [Hz]
43
44
45 Fig. 14. Variation of loss factor with frequency for flax/PP beams of different fibre
46 orientations and a fibre volume fraction of 0.31, and neat PP beams.
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 25
63
64
65
 1
2 0.1
Flax/PP, Vf = 0.40 ( = 0)
3 0.09 Flax/PP, Vf = 0.40 ( = 30)
4
0.08 Flax/PP, Vf = 0.40 ( = 45)
5
Flax/PP, Vf = 0.40 ( = 60)
6 0.07
7 Flax/PP, Vf = 0.40 ( = 90)
Loss factor

8 0.06 Neat PP

9 0.05
10
11 0.04
12 0.03
13
14 0.02
15 0.01
16
17 0
0 100 200 300 400 500 600 700 800 900 1000
18 Frequency [Hz]
19
20 Fig. 15. Variation of loss factor with frequency for flax/PP beams of different fibre
21 orientations and a fibre volume fraction of 0.40, and neat PP beams.
22
23
24 0.1 
Flax/PP, Vf = 0.50 ( = 0 )
25 
0.09 Flax/PP, Vf = 0.50 ( = 30 )
26
Flax/PP, V = 0.50 ( = 45)
27 0.08 f
Flax/PP, V = 0.50 ( = 60)
28 f
0.07 Flax/PP, V = 0.50 ( = 90)
29 f
Flax/PP, V = 0.50 ( = 30)*
Loss factor

f
30 0.06 Neat PP
31
0.05
32
33 0.04
34
0.03
35
36 0.02
37 0.01
38
39 0
0 100 200 300 400 500 600 700 800 900 1000
40 Frequency [Hz]
41
42 Fig. 16. Variation of loss factor with frequency for flax/PP beams of different fibre
43
44
orientations and a fibre volume fraction of 0.50, and neat PP beams. *Loss factors of
45 bending-twisting coupled modes.
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 26
63
64
65
 1
2 0.1
Flax/PP (Vf = 0.31) and Neat PP, L = 300 mm
3 0.09 Flax/PP (Vf = 0.31) and Neat PP, L = 350 mm
4 Flax/PP (Vf = 0.31) and Neat PP, L = 400 mm
5 0.08
6 0.07
7
Loss factor

8 0.06
9 0.05
10
11 0.04
12 0.03
13
14 0.02
15 0.01
16
17 0
0° 15° 30° 45° 60° 75° 90° PP
18 Fibre orientation
19
20 Fig. 17. Variation of loss factor with fibre orientation for flax/PP beams of a fibre
21 volume fraction of 0.31, and neat PP beams.
22
23
24 0.1
Flax/PP (Vf = 0.40) and Neat PP, L = 300 mm
25 Flax/PP (Vf = 0.40) and Neat PP, L = 350 mm
0.09
26 Flax/PP (Vf = 0.40) and Neat PP, L = 400 mm
27 0.08
28
0.07
29
Loss factor

30 0.06
31
0.05
32
33 0.04
34
0.03
35
36 0.02
37 0.01
38
39 0
0° 15° 30° 45° 60° 75° 90° PP
40 Fibre orientation
41
42 Fig. 18. Variation of loss factor with fibre orientation for flax/PP beams of a fibre
43
44
volume fraction of 0.40, and neat PP beams.
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 27
63
64
65
 1
2 0.1
Flax/PP (Vf = 0.50) and Neat PP, L = 300 mm
3 0.09 Flax/PP (Vf = 0.50) and Neat PP, L = 350 mm
4 Flax/PP (Vf = 0.50) and Neat PP, L = 400 mm
5 0.08 Coupling mode, L = 300 mm*
6 Coupling mode, L = 350 mm*
0.07
Coupling mode, L = 400 mm*
7
Loss factor

8 0.06
9 0.05
10
11 0.04
12 0.03
13
14 0.02
15 0.01
16
17 0
0° 15° 30° 45° 60° 75° 90° PP
18 Fibre orientation
19
20 Fig. 19. Variation of loss factor with fibre orientation for flax/PP beams of a fibre
21 volume fraction of 0.50, and neat PP beams. * Loss factors of bending-twisting coupled
22
23
modes.
24
25
26
27
28 0.0195
29 0.03
30 0.019
31
0.02
Damping

32
0.0185
33
34 0.01
35 0.018
36
37 0 0.0175
26.56
38 26.21 28.12
39 25.86 27.77
27.42 0.017
40 25.51 27.07
41 25.16 26.72
42 Before resonance After resonance
43
44
45 Fig. 20. Damping “carpet” plot (beam length: 300 mm, V f = 0.50,   0 , mode 1).
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 28
63
64
65
 1
2
3 0.0184
4
5 0.03
6 0.0183
7
8 0.02
Damping

0.0182
9
10
11 0.01
0.0181
12
13
0 0.018
14 166.87
15 166.52 168.44
16 166.17 168.09
167.73 0.0179
17 165.82 167.38
18 165.47 167.03
19 Before resonance After resonance
20
21 Fig. 21. Damping “carpet” plot (beam length: 300 mm, V f = 0.50,   0 , mode 2).
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62 29
63
64
65