Erosion wear assessment of sugarcane fibre reinforced polymer composites for applications of wind turbine blades

This research reports erosion wear characteristics of polyester matrix composite reinforced with short fibres of sugar cane. Erosion tests were conducted using silica sand particles of size (200–500 µm) using a locally designed air jet-type erosion test rig. The sugarcane fibre-reinforced polyester composites (5% wt.) with different fibre lengths (3, 12, 48 mm) were impacted by sand particles at a fixed velocity of 60 m/s, impact angles (30°, 60°, 90°) and erodent weight of (2500,7500) gm. ANOVA approach was used to determine the most significant parameter and their combinations that minimize the erosion rate. Scanning Electron Microscope (SEM) observations were conducted to investigate the erosion mechanisms. Results indicated that the erosion rate is greatly influenced by fibre length as well as the erodent weight. Moreover, at an erodent weight of 2500gm, the behaviour of the material turns from brittle into ductile. At an erodent weight of 7500gm, the erosion rate is inversely proportional to the fibre length.


Introduction
The necessity of using eco-friendly materials that contribute to saving the environment has encouraged researchers to develop composite materials based on natural fibres.Polymer materials reinforced with natural fibres have favourable benefits such as low density and low costs [1,2].Previous studies [3] showed a good capability of using natural fibre composites in a wide range of engineering applications.Hence, several researchers have focused on using natural fibres, owing to their physical, mechanical and tribological properties [4].There is a potential capability of using natural fibres as reinforcement materials for better resistance to harsh environments such as erosive wear conditions [5][6][7].
Erosion is undesirable and can result in problems in many engineering components.It is one of the wear modes where the damage of surfaces is caused by a repetitive impact by small solid particles carried by gas or liquid and hitting the surface.This results in the removal of material from the surfaces.In systems such as Wind Turbine Blades (WTBs) erosion is considered a big issue as it affects aerodynamic performance.This in turn encouraged scientists to work towards increasing the erosion resistance to enhance the lifetime and reduce the failures in these systems [8,9].It should be stated that erosion is one of the techniques that sometimes are implemented as desirable technique such as sandblasting to remove material.
Fibre-reinforced polymer composites have many applications in the automobile, marine, aerospace and wind turbine industries.The erosion resistance of engineering components made of these composites is of vital importance.Thus, a full understanding of erosion parameters' effects on such structural components is necessary to consider during the design of machine components [10].
Erosion is a well-known phenomenon across the industry and many efforts have been made to quantify the effect of erosion on wind turbine performance.Earlier studies [10][11][12] have been focused on the edge roughness of WTBs due to the accretion of dust and not on blade erosion.Most of these studies were qualitative with little data presented on the effect of the accretion on wind turbine performance.
Suresh et al. [13] have studied the effect of adding Arabic gum tree coal powder, Jambal tree coal powder, and Neem tree coal powders on the erosion behaviour of glass fibre-reinforced epoxy composite.Their study had a few controllable factors such as the secondary reinforcement, velocity of erodent particles, i.e., 32, 48, 52 m/sec, temperatures (40° to 60 °C) and 30°, 60° and 90° attack angles.Their results were analysed using the design of experiments to determine the best combination of parameters for minimum erosion rate.The results showed higher erosion wear rate at 30° for all the materials but Neem coal fibre gave a lower erosion rate.
The erosion characteristics of Polyetherimide (PEI) composite reinforced by carbon fabric (40% vol.) were studied by Rattan and Bijwe [14].They used solid particles of silica sand with varying attack angles at constant feed rate.The results indicated that the minimum erosion wear rate occurred at 90° attack angles despite that PEI is a polymer material whose ductility is relatively low.That is elongation to break reached 60% and maximum erosion rate occurred at attack angel of 15°-30° which confirms ductile erosion behaviour.
Another study on the erosion behaviour of epoxy polymer reinforced with a woven mat of coconut coir and short sugar cane fibre composites was conducted by Pani and Mishra [15].They used air jet erosion machine and fine silica sand particles to investigate the erosion behaviour under different speeds (48, 70, 82 m/s), and different attack angles (30°, 45°, 60°, 90°).The results showed that the type of fibre has a significant effect on erosion behaviour.Kumar et al. [7] who investigated the erosion wear behaviour of short bamboo fibre reinforced polymer composites filled with Alumina particulates.Both Pani and Mishra [15] and Kumar et.al [7] concluded that their composites behave in a semi-ductile manner.
The effect of attack angles on erosion rate and mechanical behaviour of bamboo fibres reinforced epoxy composite was studied by Gupta et al. [16].Their results showed that bamboo fibres improve the erosion wear performance of these composites.They also reported that the improvement of maximum tensile strength, flexural strength, fibre loading, impact strength and hardness were achieved at different weight ratios of bamboo fibre.In addition, the peak of erosion rate occurred at 60° and 75° angles indicating semi-brittle behaviour.
Patnaik et al. [17] implemented Taguchi's Experimental design approach on a multi-component composite system consisting of thermoplastic polyester resin reinforced with E-glass fibre and SiC particles.Their tests aimed to investigate erosion behaviour under different operating conditions.By adapting Taguchi's approach, they managed to identify the most significant factors and their interactions that affect the erosive wear.In this study [17], the authors indicated that the physical and mechanical properties of composite were improved by increasing the percentage of the fiber's contents.

(a) (b) (c)
Fig. 1 SCFP composites: a composite with 3mm chopped fibre length b composite with 12mm chopped fibre length c composite with 48mm chopped fibre length impact velocities (48-109 m/s).They found that composite behaves in brittle manner and gives maximum erosion rate at 90° and minimum rate in most cases at 30° angle.Similar results were obtained by Singh et al. [19] who implemented Taguchi method to investigate the effects of different volume fractions of sugarcane fibers (10-30 wt.%) impact velocity (30-70 m/s), impingement angles (30°-70°), standoff distance (65-85 mm) and erodent size (250-450 μm) on physical, mechanical, and erosive wear properties of epoxy composite reinforced with sugarcane fibers.In this study, the physical and mechanical properties were improved by increasing the wt.% of the fibers in the composite.The authors found that fiber wt.%, impact velocity, and impingement angles were the most significant contributing factors that influenced the response.They reported that the erosion rate was minimum at lower values of fiber content (10 wt.%), impact velocity of 30 m/s and standoff distance of 85 mm.On the other hand, when the fiber content (wt.%) and impact velocity increased, the erosion rate increased.Moreover, the results showed that the composite exhibited semi-brittle-ductile behavior whereas the maximum erosion occurred at 60° angle.
From the above survey, it seems that little work has been done on the erosion wear assessment of sugarcane fibre-reinforced polymer composites for applications of wind turbine blades.This concludes that more research on the potential of using natural fibres is needed.In this respect, studying the effect of fibre length at the lower range of fibre content on the erosion behaviour of sugarcane fibre-reinforced composites would be significant.Therefore, the objectives of this research are to study the potential of using sugarcane fibre reinforced polyester (SCFP) composite focusing on the effect of fibre length, erodent weight and impact angle on the erosion behaviour of SCFP composite.

Material fabrication and specimen preparation
The fabrication of the SCFP composite was done using the hand layup technique.The process begins with mould preparation through coating its inner surface with Vaseline to facilitate the removal of the composite.The Sugarcane Fibre (SCF) were chopped into 3 different lengths (3mm, 12mm, and 48mm).The weight of the fibre content in percentage was fixed at 5%.Three different composites corresponding to fibre lengths (3mm, 12mm, and 48mm) were fabricated.Precaution has been taken to ensure that the fibres were uniformly dispersed equally in the matrix and the variability between the three composites was minimum.The final composites are shown in Fig. 1 and their specifications are given in Table 1.Specimens were machined into plates of dimensions (25mm × 120mm × 10 mm) from each composite (3mm, 12mm, and 48mm) as shown in Fig. 2.

Experimental parameters and procedures
An air jet erosion tester is shown schematically in Fig. 3a.It was designed and fabricated locally at the British University in Egypt.The erosion testing apparatus is composed of a hopper containing the erodent sand particles, a nozzle connected to a hopper directed to a specimen holder, an air compressor with a pressure regulator, tubes and hoses connecting between the compressor and the hopper and a tank that serves as a reservoir for the sand used in the test.The apparatus was enhanced with the required facilities to test several parameters including attack angle, pressure value, impact velocity, and standoff distance.A photo of the apparatus is shown in Fig. 3b.The design of the specimen holder permits the variation of the impact angle of erodent sand particles (between 0° and 90°).More details can be found elsewhere [21].
The silica sand particles were sieved to size (200-500 µm) by using an optical microscope (OLYMPUS-BX41M-LED) shown in Fig. 4a, b.The specimens were then prepared for tests, i.e., cleaned using acetone and weighted by using high sensitivity electronic balance device (accuracy up to 4 decimal digits).Test specimens were attached to the

Experimental design
The erosion test can be conducted using a wide range of parameters, and each combination of them will yield a different response which is mainly erosion rate.Some factors were fixed at constant levels during all test runs of the experiment as shown in Table 2. Other three main factors were chosen to be controlled/varied.These controllable factors are the ones that by changing their levels, the erosion rate could be measured as a response factor.In addition, the effect of their interaction was studied and analysed using the approaches of statistical design of experiments and response surface methodology.The three controllable factors, in this case, are the impact angle, fibre length, and the weight of the solid erodent particles each at different predetermined levels, shown in Table 3.

Experimental results and discussion
Each combination of factors at their levels had three replications to ensure that each value obtained for the responses is distinguished from any outliers resulting due to undesired outer parameters.Thus, the experiment is a three-factor experiment (two varied at three levels and one at two levels) replicated three times and thus the total number of runs resulted is 54 runs in total.This is shown in Table 4 with the corresponding response value which was recorded after the experiment took place.

Analysis of erosion rate of the SCFP composites
The effects of fibre length of sugarcane (3mm, 12mm, and 48mm) and erodent weight (2500 gm and 7500 gm) on the erosion rate of SCFP composites at 30°, 60°, 90° impact angles are shown in Figs. 6 and 8. Figure 6a-c show that at an impact angle of 30° and erodent weight of 2500gm-7500gm, at 2500gm erosion rate increased slightly by 14% with increasing fibre length from 3 to 48mm but decreased significantly by 63% with increasing fibre length at 7500gm of erodent weight.
Similar results of erosion rate at 60° and 90° impact angles were produced for SCFP composite reinforced with different fibre lengths (3mm, 12mm, and 48mm) at 2500 gm and 7500gm, Figs. 6 and 7 respectively.It is clear that at 60° impact angle, Fig. 7a-c erosion rate decreases with increasing fibre length by 20% and 8% for 2500gm and 7500gm respectively.It should also be noted that insignificant effect of the erodent weight on the erosion rate for the same fibre lengths at this angle.However, at a higher impact angle of 90°, as shown in Fig. 8a-c, an increase in the fibre length showed a great reduction in erosion rate when SCFP composite impacted with 2500gm at 90° impact angle, i.e., erosion rate decreased by almost 83%.Whereas, when composite impacted with 7500gm the erosion rate decreased by 26% with increasing fibre length.
Comparison among Figs.6a, 7a, and 8a indicates that the maximum erosion rate for 3mm fibre length occurred at 90° impact angle when the composite was impacted with 2500gm erodent weight, i.e., 3.19E-06 gm.This is further reproduced in Fig. 9a which indicates a brittle behaviour.Similar behaviours were mentioned in the literature [18].On the other hand, erosion behaviour by 7500gm erodent weight gave ductile behaviour as shown in Fig. 9b with an erosion rate value of 1.03E-05 gm.
For SCFP composite reinforced with 12mm fibre length the results (Figs. 6b, 7b, and 8b) show that composite tested at 2500gm behaves in a ductile manner, i.e., maximum erosion rate obtained at 30° with erosion rate value of 3.23E-6 gm as shown in Fig. 9c.On the other hand, when the composite impacted with 7500gm of erodent weight the composite showed a ductile behaviour with maximum erosion rate at 30° (Fig. 9d).Meanwhile, erosion rate decreased to a minimum value which is  attributed to the longer fibre length 12 mm compared to 3mm.According to previously published work [20], this is attributed to a better adhesion strength between the natural fibre and polymer due to longer fibre length.
Again, the composite material reinforced with a longer fibre length of 48 mm at similar erosion conditions is shown in Figs.6c, 7c, and 8c.The material behaviour is ductile with a maximum erosion rate of 3.64E-0.6and 3.82E-06 for  2500gm and 7500gm respectively as shown in Fig. 9e, f.This result shows that at 5% wt% of the fibre in composites when increasing the fibre length, the erosion weight decreases, Moreover the behaviour of the composites proves to be mostly ductile.Which differs from the literature as the behaviour was brittle [18].On the other hand, other research [19] showed the behaviour to be semi ductile this could be due to the 5wt% that was not investigated before.It has been known that sugar cane fibres are characterised by a high adhesion strength to the polymer [20].Thus, longer fibres give more adhesion strength to the material.It was noted that little work has been published to study the synergetic effects of different wt.% and fibre lengths on the erosion behaviour of sugarcane-reinforced polyester composite.

Scanning electron microscope (SEM)
The scanning electron microscope (SEM) "Thermo Scientific Quattro ESEM" is used to examine the mechanisms of material removal by erosion action and observe the morphology of the eroded surface of the SCFP composite.the composite surfaces of 48mm fibre length (Fig. 10c).it should be emphasised that no sign of fibre debonding was detected for any of the three composite specimens tested.

ANOVA approach
The obtained data shown in Table 4 were analysed using Design Expert Software Version 11 [22] to study the influential factors affecting the response (Erosion Rate).By using the analysis of variance (ANOVA) approach, it was clear that higher-order interactions were found significant (P-values greater than 0.05) [23].The ANOVA shown in Table 5 below illustrates the different factors and interactions that constitute the total variability in the design space.Before trusting the ANOVA (Table 5), the ANOVA assumptions were analysed and revealed no violations for the assumptions.In other words, the residuals are normally and independently distributed with constant variance.Thus, the model depicted from the ANOVA is valid and can be used to develop the prediction model.The ANOVA model yielded an (R-squared) value of 0.9168, which is an acceptable value as almost 92% of the variability of the data can be explained using the developed ANOVA model.Thus, interpreting the model can be used to explain the variability within the data.The governing relation between factors at different levels and the erosion rate response is shown in the regression model obtained by Eq. ( 2).This regression model is a generic prediction for the response at any combination of the studied factors and at any level of them within the design space.experimenter and practitioner are looking for.The response surface and contour plots were investigated at the different levels of the third controllable factor erodent weight at 2500 gm and 7500gm, respectively.At 2500 gm erodent weight, Fig. 11a, b show that at the impact angle in the range between (80°-90°), and the fibre length in the range between (22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35)(36)(37), the erosion rate is minimum, whilst at the lower fibre length (3mm) and high level of attack angle (90°), the erosion rate is maximum.At 7500 gm erodent weight, Fig. 11c, d show that at the attack angle in the range between (30°-34°), and the fibre length in the range between (24-37), the erosion rate become minimum, whilst at the low level of fibre length (3mm), and low level of attack angle of (30°), the erosion is maximum.

Optimization
Design expert may give a predicted numerical value for the combination of factors that lead to maximizing or minimizing the response.The predicted model may be used to estimate the optimum fibre lengths that contribute to a minimum erosion rate at the highest erodent weight   A: Fiber Length B: impact Angle Fig. 11 Graphical representation of the erosion rate response 7500 gm and attack angle of 90°.These conditions were selected to simulate the extreme work condition for the wind turbine blades.It is predicted that the optimum fibre length of 37 mm will result in minimum erosion damage to the wind turbine blades.This output is shown in Table 6 below.

Conclusions
The erosion characteristic of sugar cane fibre reinforced polyester (SCFP) composite with different fibre lengths (3mm,12mm,48mm) were investigated under the conditions of impact angles (30°, 60°, 90°) and erodent weight particles (2500gm,7500gm) and the main conclusions may be drawn: 1. Erosion rate is greatly influenced by erodent weight, i.e., at a lower value of erodent weight (2500gm) the behaviour of the composites turns from brittle into ductile.Also, at higher erodent weight (7500gm) the behaviour of composites is ductile.2. Erosion rate of SCFP composite decreases with increasing the fibre length because of better adhesion properties.3.At 2500gm erodent weight, the changes in the erosion rate with the increase of fibre length depends on the impact angle, i.e., increasing fibre length from 3 to 48mm causes an increase in the erosion rate by 14% (at 30°) while decreases the erosion rate by 20% (at 60°) and 83% (at 90°).Meanwhile, at 7500gm erodent weight, the erosion rate decreased by 63% (at 30°), 8% (at 60°) and 26% (at 90°).4. The SEM morphologies showed that the nature of surface damage is influenced by the fibre length.That is, damage to composite surface with a fibre length of 3 mm is characterised by polymer matrix cracks and fragmentation.By contrast, damage to composite surface with a fibre length of 12 mm is characterised by small matrix fragmentation only.As to removal of the composite surface material with a fibre length of 48 mm is caused by relatively large matrix fragmentation. 5.The ANOVA showed that the three controllable factors are fibre length, impact angle, and erodent weight and their higher-order interactions are all significant.6.A mathematical Regression model with R 2 = 0.9168 which can be used to predict the erosion rate within the range limits of the controllable factor.
7. The Design Expert program predicted that using 37 mm fibre length at conditions of 90° and a higher level of erodent weight will yield minimum damage to wind turbine blade Funding Open access funding provided by The Science, Technology & Innovation Funding Authority (STDF) in cooperation with The Egyptian Knowledge Bank (EKB).

Fig. 5
Fig. 5 Fixing the SCFP Composite specimen in the sand erosion holder

Table 1
Specification of the Fiber Used in Composite Materials

Table 2
Fixed factors

Table 3
Controllable Factors

Table 4
Experimental Response Data

Table 5
ANOVA Table for the test experiment response (Erosion rate)