Use of micro-tomography for validation of method to identify interfacial shear strength from tensile tests of short regenerated cellulose fibre composites

This is an electronic reprint of the original article. This reprint may differ from the original in pagination and typographic detail. Author(s): Title: Year: Version: Please cite the original version: All material supplied via JYX is protected by copyright and other intellectual property rights, and duplication or sale of all or part of any of the repository collections is not permitted, except that material may be duplicated by you for your research use or educational purposes in electronic or print form. You must obtain permission for any other use. Electronic or print copies may not be offered, whether for sale or otherwise to anyone who is not an authorised user. Use of micro-tomography for validation of method to identify interfacial shear strength from tensile tests of short regenerated cellulose fibre composites Hajlane, A.; Miettinen, Arttu; Madsen, B.; Beauson, J.; Joffe, R. Hajlane, A., Miettinen, A., Madsen, B., Beauson, J., & Joffe, R. (2016). Use of microtomography for validation of method to identify interfacial shear strength from tensile tests of short regenerated cellulose fibre composites. In B. Madsen, A. Biel, Y. Kusano, H. Lilholt, L. Mikkelsen, L. Mishnaevsky, & B. Sørensen (Eds.), 37th Risø International Symposium on Materials Science (Article 012021). Institute of Physics Publishing Ltd.. IOP Conference Series : Materials Science and Engineering, 139. https://doi.org/10.1088/1757-899X/139/1/012021 2016 This content has been downloaded from IOPscience. Please scroll down to see the full text. Download details: IP Address: 130.234.74.31 This content was downloaded on 12/09/2016 at 05:52 Please note that terms and conditions apply. You may also be interested in: Tensile testing: a simple introduction Martin Carr Observation of iodine in oceanic plankton by scanning x-ray fluorescence micro-tomography T Ohigashi, Y Terada, A Takeuchi et al. X-ray fluorescence micro-tomography and laminography using an x-ray scanning microscope N Watanabe, M Hoshino, K Yamamoto et al. High-definition high-throughput micro-tomography at SPring-8 Kentaro Uesugi, Akihisa Takeuchi and Yoshio Suzuki Characterization of the deformation texture after tensile test and cold rolling of a Ti-6Al-4V sheet alloy B Mehdi, H Azzeddine, R Badji et al. Incorporation of measurement models in the IHCP: validation of methods for computing correction kernels J W Woolley, H B Wilson and K A Woodbury Use of micro-tomography for validation of method to identify interfacial shear strength from tensile tests of short regenerated cellulose fibre composites View the table of contents for this issue, or go to the journal homepage for more 2016 IOP Conf. Ser.: Mater. Sci. Eng. 139 012021 (http://iopscience.iop.org/1757-899X/139/1/012021) Home Search Collections Journals About Contact us My IOPscience Use of micro-tomography for validation of method to identify interfacial shear strength from tensile tests of short regenerated cellulose fibre composites A Hajlane1, A Miettinen2, B Madsen3, J Beauson3 and R Joffe1 1Luleå University of Technology, Department of Engineering Sciences and Mathematics, SE-971 87 Luleå, Sweden 2University of Jyvaskyla, Department of Physics, P.O. Box 35 (YFL) FI-40014 Jyvaskyla, Finland 3Technical University of Denmark, Department of Wind Energy, Risø Campus, Frederiksborgvej 399, 4000 Roskilde, Denmark e-mail: [email protected] Abstract. The interfacial shear strength of short regenerated cellulose fibre/polylactide composites was characterized by means of an industry-friendly adhesion test method. The interfacial shear strength was back-calculated from the experimental tensile stress-strain curves of composites by using a micro-mechanical model. The parameters characterizing the microstructure of the composites, e.g. fibre length and orientation distributions, used as input in the model were obtained by micro-tomography. The investigation was carried out on composites with untreated and surface treated fibres with various fibre weight contents (5wt%, 10wt%, and 15wt% for untreated fibres, and 15wt% for treated fibres). The properties of fibres were measured by an automated single fibre tensile test method. Based on these results, the efficiency of the fibre treatment to improve fibre/matrix adhesion is evaluated, and the applicability of the method to measure the interfacial shear strength is discussed. The results are compared with data from previous work, and with other results from the literature. 1. Introduction During recent years, cellulosic fibres (natural and manmade) have been considered as feasible replacement of synthetic reinforcement (e.g. glass fibres) for polymer composites designed for long term use in structures. The driving force for such development is the concern regarding impact of production of composites on environment, and the growing demand for sustainable and recyclable materials. This trend is also reflected in the growing number of publications on the subject as shown in literature reviews [1-2] on development of bio-based engineering materials. One of the weaknesses of cellulosic fibres is their typical poor compatibility with polymers, and the resulting weak fibre/matrix adhesion which strongly affects the mechanical properties of composite materials. This problem can be addressed by modifying the fibre surface to enhance the chemical bonding between composite constituents [3]. The problem can also be addressed by creating hierarchical reinforcement (combining nano- and microscales) with larger surface area to promote both, chemical and mechanical, fibre/matrix interaction. Hierarchical reinforcement can be created by grafting of nano-size particles on micro-size fibres, and this is believed to enhance the fibre/matrix adhesion. For example, it has been shown that the grafting of carbon nano-tubes on conventional 37th Risø International Symposium on Materials Science IOP Publishing IOP Conf. Series: Materials Science and Engineering 139 (2016) 012021 doi:10.1088/1757-899X/139/1/012021 Content from this work may be used under the terms of theCreative CommonsAttribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. Published under licence by IOP Publishing Ltd 1 fibres increases the through thickness performance of composites due to strengthening of fibre/matrix adhesion by improving the fibre surface area, and by creating mechanical interlocking [4]. In another example, composites based on hierarchical cellulosic fibres were successfully achieved by depositing bacterial cellulose on natural fibres [5]. The tensile and flexural properties of the composites of hierarchical sisal fibres embedded in acrylated epoxidised soybean oil resin were significantly improved over neat fibre reinforced composites. In the current study, cellulose nano-whiskers extracted from date palm trees are deposited onto regenerated cellulose fibres. One of the important parts of the optimization process of the fibre/matrix interface is its characterization to evaluate the efficiency of various fibre treatment methods. The usual way to characterize the fibre/matrix adhesion is by determining the interfacial shear strength (IFSS). There is number of direct and indirect methods to evaluate IFSS, ranging from tests on single fibres (fibre pullout, micro-droplet test, single fibre fragmentation) to test on full scale composites (e.g. short beam shear). However, these methods, especially those on micro-scale, are time consuming and not easy to perform, and moreover the results may significantly vary between different methods. Therefore, there is a need for a simple, fast, and yet reliable technique to measure IFSS in composites. The method used in the present study has been considered to be the most industry-friendly one [6], and is a modification of a method originally presented in [7-8]. The model scheme is based on analysis of the tensile stress-strain curves of a composite system. Two different strain values are used to obtain values for IFSS and the fibre orientation factor. The current study follows the method presented in [6-8]. All experimental parameters related to the composite morphology and fibre properties are obtained directly from experiments. This allow more reliable predictions by the micro-mechanical models since input parameters are accurately measured rather than estimated based on materials datasheets and assumptions. The output of the approach proposed in [6-8] is compared with the results obtained based on X-ray tomography data, and possible factors responsible for discrepancies between methods are discussed. 2. Theoretical background Based on the Kelly-Tyson model for the strength of aligned short fibre composites [9], and by dividing the total fibre volume fraction (Vf) into sub-fractions of fibres with given lengths (Lf), Bowyer and Bader [7] developed a model to predict stress ( σ c) as function of strain ( ε c) in composites with nonaligned fibres: cmf i j jfi fcf jfcf f ififi oc EV L dE VE d VL ε τ ε ε τ ησ )1( 4 1 )( )( )()( −+                        −+         =∑ ∑ (1) where η o is the fibre orientation factor, df is the fibre diameter, Ef is the fibre stiffness, and τ i is the interfacial shear strength. It is assumed that a critical fibre length (Lf crit ( ε )) is defined for any given strain level: i fcf critf dE L τ ε ε 2 )( = (2) Fibres shorter than Lf crit ( ε ) contribute with the index i in equation (1), and fibres longer than Lf crit ( ε ) contribute with the index j in equation (1). By selecting two strain levels from a measured stress-strain curve of a given composite system, and by following the described fitting approach in Bowyer and Bader [7], a single value for τ i (IFSS) and a single value for η o can be determined. 37th Risø International Symposium on Materials Science IOP Publishing IOP Conf. Series: Materials Science and Engineering 139 (2016) 012021 doi:10.1088/1757-899X/139/1/012021 2 It should be noted that the term Ef ⋅ε c in equations (1) and (2) corresponds to the fibre stress ( σ f). Likewise, the term Em ⋅ε c in equation (1) corresponds to the matrix stress ( σ m). This is assuming linear stress-strain behaviour of fibres and matrix. If stress-strain curves are measured for fibres and matrix, the accurate stress levels ( σ f and σ m), at a given composite strain level, can be used in the equations. 3. Experimental This section very briefly presents experimental details about the surface treatment of fibres, manufacturing of composites as well as characterization of fibre and composite properties, including detailed characteristics of the internal structure of the composite materials. More complete information concerning the applied procedures can be found in [10-13]. 3.1. Materials The regenerated cellulose fibre (RCF) Cordenka 700 Super 3, 2440 dtex, Z100 (Z100 refers to twist, 100 rev/meter) was used as received. Fibres were supplied in the form of bundles, each bundle contains 1350 fibres and average diameter of single fibre is 12.5 microns (assuming circular crosssection). Density of fibres is 1.50 g/cm3. For the modification of the fibre surface, a number of different chemicals were used. The initiator cerium ammonium nitrate (CAN), nitric acid, NaOH, MPS (organo-functional trialkoxy silane γmethacryloxypropyl trimethoxysilane) and pure ethanol (99%) were used as received without purification. Distilled water was employed as co-solvent with ethanol for the grafting reaction. Cellulose nano-whiskers (CW) were extracted from date palm tree according to the procedure reported in [11]. Polylactic acid (PLA) biopolymer was used as matrix in composites. The polymer specification is Ingeo biopolymer 2003D from NatureWorks LLC. Density of PLA is 1.24 g/cm3. 3.2. Fibre surface treatment The fibre treatment was carried out in two steps. First a silane treatment of fibres was performed and then CW were grafted onto the RCF surface (see [12] for details). The silane treatment was performed as follows: a 1600 ml mixture of ethanol and water (50/50 v%) was poured into a reactor and heated up to 65°C; after stabilizing the temperature, NaOH and nitric acid solutions were utilized to adjust pH to 7; once pH was stabilized (after approximately 2 hours), 16 grams of fibres, wound on a holder and fixed on a mechanical stirrer, were introduced into the solution (the ratio of fibres and solvent was adjusted to be 1% w/v); the polymerisation (or co-polymerisation) of MPS was carried out by adding first CAN 10-3 mol/L in the reactor and then stirring for 30 minutes. Meanwhile, the mixture was purged with nitrogen gas for 15 minutes to remove any dissolved oxygen gas. At this stage, free radicals are supposed to be created at the surface of the cellulose backbone, and are ready to react with vinyl monomers (MPS). To initiate the graft co-polymerisation, 10-3 ml/L of MPS was added to the reactor, and the nitrogen gas flow was maintained until the end of the reaction (5 hours) while stirring at medium constant rate. Fibres were then washed two times in ethanol and once in water to remove unreacted products and other impurities trapped in the fibres. The grafting process of the nano-whiskers was performed by using CW in a 0.2 wt% concentration. Distilled water (1600 ml) was poured in the reactor and pH was adjusted to be between 4 and 5. When pH was stabilized, the MPS-modified fibres were re-placed in the reactor. Prior to sonication, a solution of NaOH (0.1 M) was added to the cellulose nano-whiskers to reduce the amount of sulfonate groups and release the hydroxyl (or alcoholate) groups of CW. The CW suspension was then sonicated for 5 minutes and added into the reactor. The reaction was maintained at medium stirring rate at room temperature for 5 hours. Afterwards, the fibres were washed three times in distilled water and dried at 110°C for 4 hours under vacuum. Finally, the treated fibres were stored in a desiccator to protect the fibres from dust. 37th Risø International Symposium on Materials Science IOP Publishing IOP Conf. Series: Materials Science and Engineering 139 (2016) 012021 doi:10.1088/1757-899X/139/1/012021 3 3.3. Manufacturing of short fibre composites In order to manufacture short fibre composites with randomly oriented fibres, the continuous RCF were cut to produce 2 mm long bundles. The cutting was done by using an array of 60 razor blades (see [10] for details). The manufacturing of composite specimens consisted of two stages: 1) homogenization of RCF fibres and PLA granules to produce composite pellets; 2) injection moulding of test specimens. The homogenization of fibres and PLA was carried out in a twin-screw Brabender PLE 650 mixer. The mixing was done for 5 minutes at 215°C with screw rotation of 70 rpm. Directly after mixing, the material was pressed into flat 2 mm thick plates in a hydraulic manually operated press preheated to 80°C. The plates were cut into small pieces to produce composite pellets with a square shape (approximately 2-3 mm in size). Composite pellets were fed into a HAAKE™ MiniJet laboratory injection moulding machine (Thermo Scientific™, Karlsruhe, Germany) to produce tensile specimens according to ASTM D638 (type 5 dog-bone specimens). Detailed information about the manufacturing procedure and conditions can be found in [10]. Four different composite compositions were produced (six specimens for each): three composites of PLA with untreated fibres (5, 10 and 15 wt%), and one composite of PLA with treated fibres (15 wt%). 3.4. Characterization of fibres In order to evaluate if the fibres are degraded during the composite manufacturing process, fibre bundles were thermally conditioned to simulate the applied composite processing conditions (exposure to 220°C for 15 min). Accordingly, in total, four types of fibres were characterized: 1) untreated and unconditioned (U-UC); 2) untreated and conditioned (U-C); 3) treated and unconditioned (T-UC); 4) treated and conditioned (T-C). Tensile tests of single fibres were carried out using a Textechno machine (Favimat and Airobot2). The linear density of fibres was obtained by a vibroscopy test. The cross-sectional area of fibres was calculated from its linear density, and the density of cellulose (1.50 g/cm3). In total 50 fibres of each type were tested with gauge length of 50 mm. Tests were carried out in displacement controlled mode with a speed of 5 mm/min. 3.5. Tensile tests of composites Tensile tests of composites were carried out on an Instron 3366 electro-mechanical testing machine equipped with a 10 kN load cell and pneumatic grips. Displacement controlled mode was used with a strain rate set to approximately 1 %/min. An Instron 2620-601 extensometer with a gauge length of 12.5 mm was used to register strain. All experiments were carried out at ambient laboratory conditions (T ~ 24°C, RH ~ 42%) without prior conditioning of specimens. At least four samples for each type of composite were tested. A more detailed description of the procedure is presented in [10]. 3.6. X-ray micro-tomography X-ray micro-tomography was used to determine the fibre length and orientation distribution in the composites while the fibre orientation factor is calculated according to Krenchel’s model [14]. First, cylindrical samples of approximately 2 mm in diameter were cut from the central part of tensile test specimens of each of the four composite types. The samples were imaged using Xradia MicroCT-400 tomograph with 1.2 µm pixel size, 30 kV acceleration voltage and 3 W X-ray tube power. After reconstruction, the images were filtered with Gaussian blur (standard deviation = 1 pixel) and thresholded, yielding a binary image of fibres. The threshold value was chosen such that the average fibre diameter was 12.5 µm. Regions whose volume was less than 500 pixels were removed from analysis as they were primarily caused by imaging noise. Fibre diameter, fibre length, fibre orientation distribution as well as fibre orientation factor were then determined as described in [13]. 37th Risø International Symposium on Materials Science IOP Publishing IOP Conf. Series: Materials Science and Engineering 139 (2016) 012021 doi:10.1088/1757-899X/139/1/012021 4 4. Results and discussion This section consists of two parts; first experimental data are reported and discussed, and then the outcome of processing of results is presented with assessment of applicability of models as well as comparison with other available data. 4.1. Experimental results The typical stress-strain curves for untreated and treated conditioned fibres are presented in figure 1 (left). It is evident that the performance of the fibres is drastically affected by the treatment and conditioning. This is also shown by the analysis of fibre strength distribution as shown in figure 2 (left) by plotting probability of failure Pf as a function of stress at failure σ . The strength distribution was analysed by fitting a Weibull two-parameter statistical distribution to the experimental data (the same procedure as described in [15] is used). The parameters of the distribution (scale σ 0 and shape m) are shown in figure 2 (right). The summary of average fibre properties (diameter, stiffness, strength) along with Weibull parameters are presented in table 1. The results for unconditioned untreated fibres are in line with previously reported results for similar RCF [16-17]. The results in table 1 show that treatment reduces fibre stiffness by 5% and strength by approximately 25%. The thermal conditioning does not affect stiffness of untreated fibres and reduces it for treated fibres by about 7%. However, strength is much more affected by thermal conditioning, it is reduced by 16% for untreated fibres and by 63% for treated fibres. It also should be noted that scatter of strength has significantly increased for treated conditioned fibres (the shape parameter is much lower than for all other fibres). Since during the manufacturing of composites, fibres are subjected to the same temperatures as during the thermal conditioning, it is assumed that it has the same effect on fibre performance in composites. Therefore, in all further calculations, properties of conditioned fibres were used. The typical stress-strain curves for unreinforced PLA and the four different composites are shown in figure 1 (right). The average mechanical properties of composites are summarized in table 2. The results show that stiffness and maximum stress increases with increase of fibre content in the composites. The performance of composites with 15wt% of treated RCF is lower than that for untreated RCF composites, which is expected since fibre performance was degraded by treatment and thermal conditioning (see table 1). It should be noted that for all of the composites with untreated fibres and for unreinforced PLA, the strain at maximum stress is almost the same (between 2.6% and 2.8%). The results on mechanical performance presented here compare very well to the data for RCF/PLA composites reported in [18] (although fibre length and orientation in that study may differ). Figure 1. Typical stress-strain curves from single fibre tensile tests of untreated and treated conditioned fibres (left). Typical tensile stress-strain curves for unreinforced polymer and short fibre composites with untreated and treated fibres (right). 37th Risø International Symposium on Materials Science IOP Publishing IOP Conf. Series: Materials Science and Engineering 139 (2016) 012021 doi:10.1088/1757-899X/139/1/012021 5 Figure 2. Fibre strength distribution obtained from single fibre tensile tests (symbols in the left plot) and determination of parameters of Weibull distribution (plot on the right). The theoretical Weibull strength distributions are shown as solid lines in the plot on the left. Table 1. Single fibre tensile test results and parameters of Weibull distributions. Batch Fibres d f [µm] Strength [GPa] Stiffness [GPa] Shape parameter m Scale parameter σ 0 [GPa] <σ>a [GPa] U-UC Untreated Unconditioned 12.4 ± 0.2 0.82 ± 0.08 24.04 ± 0.39 11.84 0.86 0.82 U-C Untreated Conditioned 12.4 ± 0.2 0.69 ± 0.08 23.87 ± 0.45 14.60 0.73 0.70 T-UC Treated Unconditioned 12.4 ± 0.3 0.62 ± 0.05 22.89 ± 0.45 16.11 0.64 0.62 T-C Treated Conditioned 12.5 ± 0.2 0.23 ± 0.03 21.50 ± 0.52 8.31 0.25 0.23 aAverage fibre strength calculated by using parameters of Weibull distribution. Table 2. Average mechanical properties of neat PLA and RCF reinforced composites. Batch Stiffness a Max stress Strain at max stress [GPa] [MPa] [%] PLA 3.30 ± 0.13 54.2 ± 0.8 2.64 ± 0.12 RCF/PLA 5wt% 3.52 ± 0.20 57.7 ± 1.4 2.68 ± 0.15 RCF/PLA 10wt% 3.91 ± 0.19 63.9 ± 0.9 2.78 ± 0.15 RCF/PLA 15wt% 4.78 ± 0.13 70.2 ± 1.8 2.66 ± 0.06 Tr-RCF/PLA 15wt% 4.07 ± 0.19 52.5 ± 2.4 1.76 ± 0.10 aMeasured within strain interval of 0.05%-0.25%. The determined fibre length and orientation distributions are presented in figure 3 and figure 4, respectively. The results in figure 3 show that fibres were rather significantly reduced in length during the injection moulding process. Initially the fibre bundles were manually cut to 2 mm long pieces, and the final fibre length in the composites is within interval of 0 mm – 1 mm with majority of fibres being around 0.1 mm – 0.3 mm long. 37th Risø International Symposium on Materials Science IOP Publishing IOP Conf. Series: Materials Science and Engineering 139 (2016) 012021 doi:10.1088/1757-899X/139/1/012021 6 Figure 3. Fibre length distribution obtained from X-ray tomography of composites. Table 3 summarizes results of the X-ray tomography. It is clear from those data that all untreated fibres have very similar average fibre length (in range 0.266 mm – 0.280 mm) independent of fibre content in the composites, while treated fibres are more than two times shorter. This can be explained by much lower mechanical properties of the treated fibres due to the severe thermal degradation which makes it much easier to break up fibres during composite manufacturing. The fibre orientation distribution in figure 4 show that fibres are not completely randomly oriented and there is some fibre alignment towards the loading direction (majority of fibres are oriented within 0º-45º with respect to the loading direction). However, the alignment with loading direction is not dominant. It is also worth noticing that treated fibres are less oriented than untreated fibres, which is likely related to their length being shorter. Figure 4. Distribution of angles between fibre direction and loading direction, obtained from X-ray tomography of composites. 37th Risø International Symposium on Materials Science IOP Publishing IOP Conf. Series: Materials Science and Engineering 139 (2016) 012021 doi:10.1088/1757-899X/139/1/012021 7