How the Test Setup Can Affect Single Fiber Tensile Testing

ABSTRACT

Tensile testing is the most common method to investigate natural fibers. The fibers’ mechanical behavior can be considered non-linear and is influenced by viscoelasticity, plasticity, and environmental conditions. Very often such fibers are tested by gluing them with an adhesive onto a sample holder. Such a system consisting of a polymeric sample holder, adhesive, and natural fiber is complex and there is a risk that the mechanical response measured is a mix of the different contributions of those components. In this work, the key components for tensile testing of natural fibers ‒ sample holder and adhesive ‒ are investigated, to determine their influence on the measurement results. In order to isolate the influence of the measurement setup, the natural fiber is replaced with a platinum wire, which is purely linear-elastic. Hence all non-linear contributions from sample holder or adhesive can be identified. The main influence factor on the results was the glue used for fixating the fiber on the sample holder. Epoxy resin was found to be best suited. Taking these findings into account, a series of tensile tests was performed on cellulose-based natural fibers for demonstration but is applicable to any natural fiber.

摘要

拉伸试验是研究天然纤维最常见的方法. 纤维的力学行为可以被认为是非线性的，并受到粘弹性、塑性和环境条件的影响. 这种纤维通常通过用粘合剂将它们粘合到样品支架上来进行测试. 这种由聚合物样品支架、粘合剂和天然纤维组成的系统是复杂的，并且存在测量的机械响应是这些成分的不同贡献的混合的风险. 在这项工作中，研究了天然纤维拉伸测试的关键部件——样品支架和粘合剂，以确定它们对测量结果的影响. 为了隔离测量设置的影响，天然纤维被纯线性弹性的铂丝取代. 因此，可以识别来自样品支架或粘合剂的所有非线性. 影响结果的主要因素是用于将纤维固定在样品支架上的胶水. 发现环氧树脂最适合. 考虑到这些发现，对纤维素基天然纤维进行了一系列拉伸试验以进行证明，但适用于任何天然纤维.

KEYWORDS:

• Cellulosic fibers
• tensile testing
• validation
• mechanical properties

关键词:

• 纤维素纤维
• 拉伸试验
• 验证
• 机械性能

Introduction

Cellulose-based fibers are abundant in nature and a main component of many products in the bio-based industrial sector. For the production of bio-based composites, fibers like flax, hemp, kenaf, jute and ramie are interesting (Maity, Prasad Gon, and Paul 2014; Malik et al. 2022; Pickering, Efendy, and Le 2016) and for papermaking as well as for the production of man-made fibers, the processing of wood fibers is crucial as well. Furthermore, due to varying regional availabilities the focus on natural fibers for bio-based composites is different around the world ranging from bagasse and banana fibers to pineapple and sisal fibers. However, there is no debate that tensile testing is the most established technique to test the mechanical behavior of all these different fiber types (Baley 2002; Chokshi et al. 2022; Horbelt et al. 2021; Jajcinovic et al. 2018,; Page, El-Hosseiny, and Winkler 1971; Placet, Cissé, and Lamine Boubakar 2014; Venugopal and Boominathan 2022).

The handling of the fibers during testing can vary due to their geometrical dimensions and one can differentiate for the testing procedure between two different approaches: (i) testing the fibers directly or (ii) fixing the fibers with adhesive onto a sample holder. Lots of natural fibers such as flax, hemp, sisal, coir, and jute (Pickering, Efendy, and Le 2016) usually have a length which is long enough to directly clamp these fibers into the tensile testing device. However, very often the clamping is not working properly and slippage or breakage at the clamping points of the fiber can occur. This makes the evaluation of such measurements questionable. Furthermore, there are other natural fibers, e.g., bamboo (Zhang et al. 2018) or wood-based fibers (Horbelt et al. 2021; Jajcinovic et al. 2018), which have a limited fiber length of only 2–5 mm. This limit requires the utilization of a sample holder (SH), which translates the applied force from the tensile tester clamp into the fiber, therefore, the fiber is glued with an adhesive onto an SH, which is made from a polymeric material such as acrylic glass or paper.

Interestingly, in literature one can find many examples of the tensile testing approach with an SH – even for longer fibers such as whole bamboo strips (Salih, Zulkifli, and Husna Azhari 2022) and single flax fibers (Amroune et al. 2020; Charlet et al. 2007). However, there is no thorough investigation available to check if by using such sample holder systems the adhesive as well as the polymeric SH can contribute to the material response during deformation. As a result, this could lead – especially for shorter fibers – to a wrong assessment of the resulting a force-displacement curve which could be a mix of contributions from the natural fiber, the adhesive, and the polymeric SH.

In this work, we want to investigate the sample holder system consisting of polymeric SH, adhesive and natural fiber to discriminate between the influencing factors and provide some good practices for the testing of natural fibers within such sample holder systems.

To further illustrate this, let’s consider a simple example of a dynamic or cyclic tensile test performed on a material as presented in Figure 1(a). In this work, we will use such a measurement protocol to investigate our materials. If, for example, a force-controlled dynamic experiment is performed to study the mechanical performance of a fiber, a step-wise increasing cyclic force is applied to the material with a short holding time (conditioning) in between the individual steps and the displacement signal is recorded. If the material is linear elastic, the response of force and displacement over time would look as illustrated in Figure 1(b) – the loading and unloading curves are all overlapping for each circle. Here, one can use a simple spring to describe the linear elastic behavior of this material. However, this is an ideal representation and not applicable to cellulose-based natural fibers. They are known to exhibit a non-linear response with viscoelasticity and plasticity playing a role, which can be further affected by humidity and temperature. Therefore, the result of a cyclic tensile testing is looking rather as illustrated in Figure 1(c) with a non-linear response between the loading and unloading curves in each cycle. Here, the contributions of elasticity, viscoelasticity and plasticity are illustrated by a simplified combination of spring, dashpot, and sliding frictional elements.

Figure 1. (a) Force-controlled measurement with the displacement closely following the behavior of the force; (b) force-displacement curve displaying the ideal behavior of a linear-elastic material; (c) real force-displacement curves of a natural fiber with non-linearity and plasticity.

If we consider now a testing procedure which has the single fiber fixed onto a polymeric SH with an adhesive, the situation gets even more complicated. As illustrated in Figure 2(a), the main contributors to the deformation can be the polymeric SH, the adhesive, and the single fiber. However, within the testing conditions it is not possible to distinguish between the different contributions and, therefore, the resulting force-displacement data is a mixed signal of all three components. In this work, we propose a simplified approach and a few experiments to validate the possible influence factors on the testing of single fibers within a polymeric SH.

Figure 2. (a) In a cyclic tensile experiment of a natural fiber which is fixed on a sample holder (SH) with an adhesive, it is not possible to clearly distinguish the natural fiber’s mechanical response because it can be a mix of contributions of all components. (b) The sample holder system can be simplified by exchanging the fiber with a Pt-wire. This way, one of the three possible viscoelastic/plastic systems is replaced with a linear-elastic system with known properties. In the test setup with the platinum wire all non-linear response and plasticity of the system is caused by polymeric SH or adhesive.

As a first step, which is illustrated in Figure 2(b), the natural fiber – our material of interest – is replaced by a linear-elastic material with well-characterized properties: a Platinum (Pt)-wire. The Pt-wire is strictly linear elastic and has a well-known stiffness, therefore, it reduces the complexity and by testing its mechanical response, we can evaluate the non-linearity (viscoelasticity and plasticity) originating solely from the polymeric SH and adhesive. Furthermore, by testing such a stiff material as Pt within the polymeric SH, a possible stiffness reduction induced by the SH and/or adhesive can be detected by referring to literature values for Pt-alloys.

As a second step, the SH is investigated as a blank (without the fiber and the adhesive) to account for any effects which could be related to non-sufficient equilibration time of the polymeric SH. Next the adhesives are characterized in two ways: (1) the mechanical strength of different adhesives is investigated in a controlled environment with the Pt-wire, and (2) the wetting behavior of selected adhesives on natural fibers is observed in a scanning electron microscope (SEM). After the method was validated and experimental settings for both SH and adhesive were improved, the whole system was tested with industrial processed wood fibers.

The described experiments can be regarded as a validation procedure for natural fibers which are tested for their mechanical behavior within a sample holder system (polymeric SH + adhesive + natural fiber) to make sure that the SH and adhesive do not affect the results of the single fiber tensile test.

Materials and methods

All experiments were performed at controlled environmental conditions with a temperature of 23°C and a relative humidity (RH) level of 50%.

Sample preparation

The samples used in the experiment were platinum wires, glued to the sample holder. The platinum wire (Pt-wire) was made from PtNi 90/10 alloy (tensile strength (TS) = 2231 N mm⁻², breaking force = 0.71 N), manufactured and provided (including reference data) by Carl Haas GmbH spiral spring factory (Germany). The Pt-wire was 0.5 mm in width and 0.05 mm in thickness. The sample holder (SH) was laser-cut from a 0.3 mm thick sheet from Poly(methyl methacrylate), which is also called PMMA, or acrylic glass (TS = 75 N mm⁻²; E = 3.30 GPa), provided by Topacryl AG (Switzerland). The adhesives used to glue the fibers onto the sample holder were nail polish (NP), UHU 4557 super glue (SG), 2-component epoxy resin UHU Plus Endfest (EP) and a Cold Weld two-part epoxy system from J-B Weld (JB).

Similarly, as longer fibers are commonly fixated in a cardboard or paper frame (Loganathan et al. 2021; Subramanya et al. 2017), the short wood pulp fibers are fixated with an adhesive onto a polymeric SH, as illustrated in Figure 3.

Figure 3. (a) Pt-wire glued on the sample holder and vertically mounted in the DMA device and (b-c) schematic depiction of the sample setup (b) before the measurement with still present connecting bridges and (c) during the measurement after melting of the connecting bridges.

To stay in the same dimensional scale as the measured fibers, the Pt-wires used in the validation were cut to a length of 2–3 mm. The final samples were then prepared as follows. First, a droplet of the adhesive was placed on each side of the gap on the SH and left for approximately half a minute to cure. Then, a Pt-wire was placed across the gap, so that both ends were positioned on the adhesive. Next, another droplet of the adhesive was placed on top of the previous droplet, fully immersing both ends of the Pt-wire. The whole setup was then left to cure under constant conditions (T = 25°C; RH = 50%). Here, the curing time was depending on the investigation. For comparison of different adhesives, the curing time was 24 h, while for the curing investigation, the curing times were 24 h, 48 h, 72 h and 96 h respectively.

The industrial wood fibers tested for demonstration were prepared with EP glue, an adhesive curing time of 48 h, at 23°C and 50% RH. They were measured with the same protocol as used in the Pt-wire investigation (Figure 4b) and an individual force-controlled protocol was created just to showcase the effects of viscoelasticity and plasticity.

Figure 4. (a) Illustration of the direct mounting of the Pt-wire to the DMA device and (b) cyclic displacement-controlled measurement protocol for the validation of the Pt-wire. Ten different displacement rates were applied with a two-minute relaxation time between the ramps. The maximum displacement was kept constant at 15 µm for each cycle.

Dynamic mechanical analysis – DMA and data analysis

Tensile testing was performed by using a Dynamic Measurement Analysis (DMA) device (DMA850, TA Instruments, USA). The sample system was positioned vertically, as it is displayed in Figure 3(a).

Initially the position of the mounting clamps was locked and prior to the start of the measurement, the connecting bridges (Figure 3) were melted away using a soldering rod with a narrow and flat tip, so both parts of the SH remained connected only by the sample (i.e. the Pt-wire). During the measurement, the force was applied vertically, compare Figure 3(c). The raw data was first analyzed in Trios, a software provided by TA Instruments for the DMA device and later in MS Excel.

Platinum wire investigation

First, the mechanical properties of the platinum wire were investigated. The wire was mounted directly into the DMA without an SH (Figure 4(a)) to determine its force-displacement. Slippage can be excluded as the wire was wrapped around the clamp several times. For the DMA analysis, a displacement-controlled measurement protocol consisting of 10 cycles with constant maximum displacement L_max = 15 µm and strain rates r increasing from 0.113% s⁻¹ to 800% s⁻¹ (Figure 4(b)) was designed to confirm the strictly linear elastic material behavior. A conditioning step of two minutes was implemented to ensure that the Pt-wire has enough time to fully relax before the next cycle of measurement starts.

After the tensile testing was complete, the cross-section of the Pt-wire was determined by microtome cutting. Each Pt-wire was embedded in a glycol methacrylate resin and the resin was left for 24 h to cure. The embedded wire was cut at a minimum of four different positions along the length. For each of those cuts, an optical image of the cross-section was recorded. The outlines of the cross-sectional area of the wire (Figure 5(a)) were then drawn manually in an image analysis software. Afterward, the image was binarized as illustrated in Figure 5(b).

Figure 5. Image analysis of the Pt-wire after the microtome cutting: (a) microscope image and (b) binarized cross-section of the wire.

Following the binarization, a Matlab script was used to calculate the cross-section area of the wire. Several different cross sections of the same wire were recorded in our microscope imaging setup (Lorbach et al. 2012; Wiltsche et al. 2011). This way, the mean area of the Pt-wire cross-section was determined. The mean value of the cross-sectional area of the Pt-wire was determined to be (310 ± 20) µm². The resulting force vs. displacement curves were then corrected by being shifted, so that the starting point of each curve was set to 0 mm and 0 N. The strain ε was calculated from the shifted displacement data, considering the initial length of the Pt-wire. The stress σ was calculated from the shifted force data, considering the cross-sectional area of the Pt-wire. The stress-strain curve was plotted for each strain rate and the rate-dependent longitudinal modulus E_r was determined as the slope of the linear part of the curve.

Sample holder investigation

A sample holder investigation was performed to determine the effect of the sample holder on the measured results. For this purpose, a single blank sample holder was mounted into the DMA without melting the connecting bridges. Repeated measurements were made using a force-controlled measurement protocol, immediately followed by a displacement-controlled measurement protocol, see Figure 6. Five such measurements were carried out over the course of one day, with four hours of pause between measurements. A more detailed description of the different steps in the measurement protocol is provided in section S1 (Figure S1) in the ESI.

Figure 6. (a) Force-controlled and (b) displacement-controlled measurement protocols for investigation of the blank sample holder.

Adhesive investigation

After evaluation of the mechanical properties of the Pt-wire and determining the effect of the sample holder, different kinds of adhesives were investigated, determining their suitability and the ideal curing time for the most promising adhesive.

Adhesive comparison

The Pt-wires were glued with four different adhesives: nail polish (NP), super glue (SG), 2-component epoxy resin (EP) and a Cold Weld two-part epoxy system J-B Weld (JB) before tensile testing with a force-controlled measurement protocol which included a preconditioning time and an increasing maximum force in each cycle (Figure 7(a)). A more detailed description of the different steps in the measurement protocol is provided in section S2 (Figure S3) in the ESI.

Figure 7. (a) Preconditioning and the first four cycles of the force-controlled measuring protocol for the adhesive comparison. The maximum force is steadily increased until 360 mN. (b) Preconditioning and the first four cycles of the force-controlled measuring protocol for the curing time investigation.

Adhesive curing time

Four Pt-wires were glued with EP adhesive and measured 24 h, 48 h, 72 h, and 96 h after preparation. The measuring protocol used for the measurements was a modified version of the protocol used in the adhesive comparison, where a 30 s conditioning step (conditioning I) was added at the maximal applied peak force F_max in each cycle, as displayed in Figure 7(b), in order to check the behavior of the adhesive under constant load.

Results & discussion

Pt-wire investigation

For the course of the validation study, single pulp fibers were replaced with linear-elastic platinum wires with well-known mechanical properties. The Pt-wires were mounted directly into the DMA device (Figure 4(a)) before tensile tests with 10 increasing strain rates r ranging from 0.113%s⁻¹ to 800%s⁻¹ were performed on the same wire.

The force vs. displacement curves for each rate are displayed in Figure 8(a). They overlap at all strain rates with the individual slopes becoming steeper and constant after the force reaches 0.01 N. Except for the initial part of the curves, which is caused by an initial sagging of the wire between the clamps, the Pt-wire is behaving linear-elastic, with no visible influence of the strain rate. By plotting the determined E_r against the logarithmic value of the corresponding strain rate log r (Figure 8(b)), no significant change in modulus can be observed at different r.

Figure 8. (a) Corrected force-displacement curves for each of the applied strain rates r and (b) the change of the mean rate-dependent modulus E_r in dependence of the displacement rate (logarithm of r). The error bars represent the confidence interval (Student distribution; α = 0.05; N = 4).

The value of the measured modulus E_r is (159 ± 3) GPa. The error bars give the confidence interval (Student’s distribution; α = 0.05; N = 4). Since the Pt-wire is actually a PtNi 90/10 alloy, the measured value is slightly higher than E_Pt (146.9 GPa) and significantly lower than E_Ni (220.7 GPa) (Wilson 2005).

Sample holder investigation

To determine the influence of the sample holder on the results of single fiber tensile testing, four force-controlled (Figure 6(a)) tensile tests were performed on a blank sample holder, in four-hour-intervals. Each force-controlled measurement was directly followed by a displacement-controlled measurement (Figure 6(b)). When comparing the corrected results of the force-controlled measurement, a shrinkage of the sample holder can be observed throughout the first measurement (Figure 9(a)), resulting in a total shrinkage of 2 µm. Considering a free span length for the fibers during the measurement of 0.8–0.9 mm, this shrinkage effect accounts for 0.25% of the fiber length. This introduces an artefact to measurements, which particularly hides plastic deformations at small loads.

Figure 9. Force vs. displacement curves of force-controlled measurements of a blank sample holder, (a) initial measurement, (b) measurement after 4 h, (c) measurement after 8 h and (d) measurement after 12 h.

In the second measurement, 4 h after the start of the first measurement (Figure 9(b)), the shrinkage is still present, but smaller (close to 0.4 µm). With increasing time, one can see that the shrinkage is further decreasing until 12 h after the start of the initial measurement (Figure 9(d)), it is so small that it can be completely neglected. The results of the displacement-controlled measurements which are presented in Figure S2 in the Electronic Supplementary Information (ESI) indicate the same findings. The force-controlled experiments showed the effect of shrinkage very clearly and are easier to interpret, therefore, all subsequent experiments are also performed force-controlled. Since the sample holder material PMMA is hygroscopic, the shrinkage can be attributed to the change in relative humidity in the measurement instrument compared to the ambient conditions. Hence, enough equilibration time has to be considered if a hygroscopic material is chosen as sample holder and changes in relative humidity are occurring in the measurement setup.

Adhesive investigation

Influence of adhesives on mechanical strength

In general, one can expect a stronger interaction between the adhesive and the natural fiber because the fiber, the adhesive as well as the acrylic glass SH are polymeric. Furthermore, the fiber is porous and more compliant compared to the Pt-wire. Therefore, we used the Pt-wire as a test for the adhesive because if it works without slippage for the wire, it will for sure work with the natural fibers. In this investigation, Pt-wires were glued on SHs with four different adhesives: nail polish (NP), super glue (SG), epoxy glue (EP) and JB-weld (JB). Ambient conditions were kept constant (T = 23°C, RH = 50%). After a curing time of 48 h, cyclic tensile tests were performed on all the samples, following the protocol displayed in Figure 7(a). The results for each adhesive were analyzed and are presented in Figure 10.

Figure 10. Comparison of the results in adhesion investigation: (a) nail polish (NP), (b) super glue (SG), (c) two component epoxy glue (EP) and (d) cold weld two-part epoxy system J-B weld (JB).

The samples glued with NP (Figure 10(a)) display a high degree of slippage which increases with every following cycle until the critical failure occurs. The critical failure did not happen due to breakage of the Pt-wire, but because the Pt-wire was pulled out of the adhesive. In contrast to this extreme case, the samples glued with SG (Figure 10(b)) and EP (Figure 10(c)) exhibit a more promising behavior. Although both force-displacement curves are slightly bent at smaller forces, which could be due to a sag of the Pt-wire which was already present in the measurements in Figure 8, all the loading and unloading curves are positioned almost on top of each other, which is implying the absence of any slippage. Finally, JB (Figure 10(d)) displays even better performance for the loading and unloading curves, since the curves are positioned exactly on top of each other and are nearly linear. The total displacement is significantly lower in comparison to the other adhesives, which is very close to the behavior of the Pt-wire, when directly mounted to the DMA. However, JB, in comparison to other adhesives, peels quite easily from the surface of the PMMA sample holders. This is an issue at higher applied forces and causes the breakage of the glue from the SH, which is limiting the applicability of JB, so we excluded JB.

After performing these tensile tests on Pt-wires fixed with four different kinds of adhesives to the SH, only SG and EP were chosen for further testing.

During these tests, the effect of adhesive on the tensile behavior of the Pt-wire was clearly demonstrated. By choosing an unsuitable adhesive, one can seriously distort the fiber’s response as demonstrated with the linear elastic Pt-wire. It is recommended to test the adhesive on a linear elastic material before applying it for the tensile testing of natural fibers.

Wetting behavior of SG and EP adhesives on natural fibers

In the next step of our investigation, we looked into the possible wetting and penetration of the glues into the fiber, which could also influence the results of the measurements (Fischer et al. 2014). For this purpose, SEM images of single pulp fibers, glued with SG and EP adhesives were recorded at different magnifications as presented in Figure 11.

Figure 11. Comparison of SEM images for SG at (a) 90x, (b) 1000x and (c) 3000x magnification and EP at (d) 90x, (e) 250x and (f) 750x magnification.

Even at the lowest magnification (90×), the difference between the wetting behavior of the SG and EP is obvious. While SG (Figure 11(a)) seems to create a rough surface of the gluing spot, maybe indicating a chemical reaction, the surface of EP (Figure 11(d)) is smooth. The epoxy material visible on the fiber was caused by handling during the gluing process. At higher magnifications, wetting of the fiber can be observed on the SG sample (Figure 11(b), while the surface of the EP sample (Figure 11(e)) stays smooth. Already previous investigations indicated that super glue is wetting the fiber easily (Fischer et al. 2014). In the zoom-ins (Figure 11(c,f)), the difference in wetting behavior is clearly visible.

Therefore, also the SG adhesive was excluded from the investigation. EP remained as the adhesive with the most promising performance. An investigation on the curing time of the EP adhesive, which is presented in section S2.2 and Figure S4 in the ESI, shows that curing is finished after 48 hours, and the sample holder system is stable after that.

Natural fiber testing

To demonstrate that the validated sample holder system works also for single natural fibers instead of Pt-wires, industrially processed wood fibers (chemi-thermomechanical softwood pulp fibers, CTMP) were chosen as representative test specimens for cellulose-based natural fibers. They were prepared in accordance with the results of the validation investigations (EP glue; curing time 48 h, at 23°C and 50% RH) and measured with the same protocol as for the Pt-wire investigation (Figure 4(b)). In Figure 12, the results for the CTMP fiber is presented.

Figure 12. (a) Change of the modulus E_r in dependence of the logarithmic displacement rate log r for four different industrial wood fibers (CTMP) (values are mean ± standard deviation). (b) Displacement response of a force-controlled protocol for a combined viscoelastic and plastic study on three different CTMP fibers. In (c) a zoom-in into the viscoelastic region of the protocol (first 14 min) is provided.

While there is no significant change in the E_r of Pt-wires at different displacement rates r (Figure 8(b)), CTMP shows a significant rate dependency of E_r, as displayed in Figure 12(a). A more comprehensive study obtained for different cellulosic fibers confirms these results and revealed a strain rate dependence of up to 20% per decade (Zizek, Czibula, and Hirn 2022). As the system has been successfully validated with linear elastic Pt-wires it can be concluded that the rate dependency of E_r, and the observed plastic deformation is indeed originating from the fibers. In Figure 12(b), the complexity of the applied force-controlled protocol is increased by combining a viscoelastic regime followed by a cyclic regime to investigate plasticity. In the displacement response a clear difference in the response of the fibers is visible, especially in the plastic behavior. Here, the fibers fail at different force levels (75 mN to 200 mN). The zoom-in (Figure 12(c)) into the viscoelastic region also exhibits differences in viscoelastic deformation.

Conclusions

In this work, the main components of the sample holder system for the testing of natural cellulosic fibers ‒ polymeric sample holder (SH) and adhesive ‒ were investigated to determine their influence on the results of single fiber tensile testing. The sample holder system – which consists of the polymeric SH, the adhesive, and the natural fiber – was simplified by replacing the natural fiber with a platinum wire. Such a metallic wire is purely linear-elastic, and not sensitive to changes in ambient humidity. Hence all effects indicating viscoelasticity, plasticity or hygro-expansion can be related to the SH or the adhesive. This way, several influencing factors were pinpointed and need to be considered when designing an experimental procedure for the testing of single natural fibers.

First, it was found that if a hygroscopic material is used for the SH – such as acrylic glass in this work – sufficient time for equilibration must be provided, otherwise shrinkage or expansion of the sample holder was found to introduce measurement artefacts indicating apparent plastic deformation.

Next, the tensile strength performance of four different adhesives was investigated: nail polish (NP), super glue (SG), two component epoxy resin (EP) and a Cold Weld two-part epoxy system (JB). NP performed worst, strong slippage of the fibers in the gluing was observed, indicating a plastic deformation of the fibers in the measurement which actually never took place. JB has excellent linear elastic behavior, but cannot be used for higher loads because it fails by detaching from the PMMA sample holder. This investigation proved that the adhesive is a very important component that needs careful consideration and needs to be checked to avoid any contribution on the material response of the fiber.

Since the performance of SG and EP in the tensile testing was good, SEM imaging was performed on both adhesives. The imaging confirmed the wetting of the fibers by SG as reported in literature (Fischer et al. 2014). Therefore, SG was eliminated, and EP was chosen for further investigations. The curing time of an EP glue on an acrylic glass sample holder was found to be 48 h, before that elevated plasticity was observed.

To demonstrate that the validated system is applicable for tensile testing of natural fibers, industrial wood fibers were tested in accordance with the validation results. These natural fibers exhibited a clear strain rate dependency of the modulus and also pronounced viscoelasticity and plasticity was observed. Due to the validation procedure applied it can be excluded that these effects descend from the measurement setup.

Finally, the findings and experimental procedures in this work can be applied to the testing of any natural fiber which is fixed onto an SH with an adhesive.

Author’s contribution

M.Z. and C.C. wrote the main manuscript. M.Z., C.C. and U.H. outlined the experimental work. M.Z. performed most of the experiments and analyzed the data. Single fiber testing demonstrations were performed and analyzed by C.C. U.H. provided funding and supervised the work. All authors reviewed the manuscript and contributed to the interpretation of the results.

Consent

The work presented in this manuscript does not raise any issues concerning consent.

Ethical approval

The work presented in this manuscript does not raise any ethical issues.

Acknowledgments

The financial support by the Austrian Federal Ministry for Digital and Economic Affairs and the National Foundation for Research Technology and Development is gratefully acknowledged. We also thank our industrial partners Mondi Group, Canon Production Printing, Kelheim Fibres GmbH, SIG Combibloc Group AG for fruitful discussions and their financial support. Special thanks to Angela Wolfbauer of the Institute of Bioproducts and Paper Technology, Graz University of Technology for sample preparation and Dr. August Brandberg for helpful suggestions.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author, U. H., upon reasonable request.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15440478.2024.2328264

Additional information

Funding

M. Z. and U. H. thank the EU Horizon 2020 Program under Marie Sklodowska-Curie Grant Agreement 764713, ITN Project FibreNet. C. C. acknowledges the Hertha Firnberg program [project no. T 1314-N] of the Austrian Science Fund (FWF) for funding.