Influence of water on the short and long term mechanical behaviour of polyamide 6 (nylon) fibres and yarns
Polyamide fibres are known to be sensitive to water, and this has limited their marine applications. However, their low stiffness and high strength could be attractive for shallow water mooring line ropes, particularly for floating wind turbines. Such applications require the influence of water on mechanical behaviour to be fully understood, and this paper presents results describing how water affects static and creep response of polyamide 6 fibre yarns. First, the effect of water on fibre Tg is quantified, then tensile behaviour is examined. Finally, a model based on Schapery’s non-linear creep analysis is identified for different humidity conditions. The results show that while moisture has a small effect on short-term tensile behaviour, long term response is strongly affected and this must be considered in any marine application.
KeywordsAging Moisture Fibre Creep Viscoelastic model
Polyamide fibres are used very extensively in a range of industrial applications from textile fabrics to climbing ropes, tire cords and architectural structures. They were first marketed in the 1930s so there is considerable experience available; recent overviews of their chemistry, manufacture and properties can be found in Deopura et al. (2008) and Bunsell (2009). There is also an increasing interest in these fibres for marine applications. Nylon fishing nets and single point offshore moorings are well established, but the former involve intermittent immersion while the latter are usually replaced after 1 or 2 years. Today nylon ropes are being considered for station keeping of ocean energy structures, and in particular for floating wind platforms (Ridge et al. 2010; Weller et al. 2015). This is an application where long term durability is critical, and the combination of high loads and continuous immersion is encouraging new developments in fibre coatings and improved rope constructions. There is therefore renewed interest in the influence of water on nylon properties, particularly those of polyamide 6 fibres.
If we first consider the microstructure of nylon fibres and their influence on tensile behaviour; Marcellan et al. have provided a detailed description of the micro-mechanisms which contribute to nylon 66 fibre deformation, and which produce a non-linear stress–strain response (Marcellan et al. 2006). Initially isotropic amorphous regions are loaded but as load is increased the oriented amorphous and crystalline regions are progressively elongated, leading to stiffening. In the final stage of loading, which can be considered as viscoplastic, macrofibril sliding results in damage.
Various authors have examined the influence of water on bulk nylon properties. For example, Arhant et al. presented a water diffusion model which accounted for the change in water uptake due to the passage from the glassy to the rubbery state during wet aging (Arhant et al. 2016). Experimental examination of the rates of transport of liquid water along a wide variety of yarns supports the view that water travel occurs essentially by capillary action (Hollies et al. 1957).
Hunt and Darlington studied the creep behaviour of bulk nylon 6.6 and quantified the influence of moisture (Hunt and Darlington 1978, 1979, 1980). They found the creep behaviour under varying humidity could be estimated using creep data at constant stress and humidity.
There have been various attempts to represent the mechanical response of polymer fibres. Militky (2009) proposes the following four categories: continuum, micromechanical constitutive, structural, and multi-phase models. These differ in the way in which the fibre microstructure is accounted for. The continuum models ignore the molecular aspects of polymers and include elastic and viscoelastic behaviour of a homogeneous body. One example of this type of approach was proposed by Chailleux and Davies, who represented the time dependent strain in polyester fibres (Chailleux and Davies 2005). A certain number of identification steps, based on creep and recovery curves and specific load-unload sequences, are required to predict strain as a function of loading history. Once these have been performed the response to more complex loading sequences can be predicted. The use of this type of model to quantify moisture effects is discussed further below. Various spring and dashpot representations have been applied in other studies to account for the time dependent response. Micromechanical constitutive models can also use springs and dashpots, but these are associated with physical elements of the fibre such as molecular entanglements and crystallites. An example is the work of Fotheringham and Cherry to predict yield stresses (Fotheringham and Cherry 1978), with an Eyring activation energy model to account for simultaneous cooperative movement of polymer chain segments. Structural models are based on calculation of the macroscopic mechanical behaviour based on the responses of the structural elements. Many consider the fibre as a composite material, including polymer fibrils in a less ordered matrix, such as the series model of Northolt et al. (1995). Finally, the multiphase models assume that the different phases of a fibre are continuous; the Takaynagi model is the best known (Takayanagi et al. 1967). This is based on combining the contributions of the amorphous phase, the crystalline phase and a tie chain phase. The behaviour of each of these phases can be defined in terms of degree of crystallinity and orientation, but they provide a very simplified approach to properties.
This paper will focus on the physical influence of water on mechanical properties of polyamide 6 fibre yarns; hydrolysis is not considered here. First, materials and methods specially developed for the study will be exposed. Second, the water ingress and the mechanical properties of dry fibres will be examined in order to define reference behaviour. Finally, the mechanical properties will be defined in different humid environments in order to quantify the influence of the amount of water in the fibre material. The overall aim of the work is to provide the tools for a multi-scale durability analysis of fibrous structures such as nets and ropes.
2 Materials and methods
In this section we present the material used in the present study and its general physical properties.
In addition, the methods used for the investigation are reported. Some of these techniques have been specially developed for this study, so in these cases, experimental tests and identification methods are described in specific sub-sections.
The study is focused on polyamide fibres, one of the most widely-used polymers on account of its ductility and high strength. Nylon fibres (polyamide) are produced from amide groups (–CONH–); there are several different types of nylon fibres: polyamide 6.6/6.12/4.6/6/12 (the number refers to the carbon atoms around the amide group in the monomer), which have found a wide range of applications. The fibre used in the present study is a nylon 6, grade 371 manufactured by Nexis™ fibres, and has the reference 1880 f 280, with a linear weight of 190.5 tex (g/km).
The theoretical enthalpy of fusion of 100% crystalline polyamide 6 was taken as 188 J/g [value from Arhant et al. (2016)]. Using this value and the ΔHf obtained by DSC, a degree of crystallinity of 40.1% was determined.
2.4 Glass transition temperature measurements
Then tests were performed under different environmental conditions inside a temperature and humidity controlled chamber. This was filled with silica gel to get the properties at 0%RH and then different humidity conditions were used to characterize the influence of water content on the properties. Specimens were tested at a frequency of 1 Hz and a dynamic strain of 5 × 10−3. The tests were conducted from 25 to 150 °C at 2 °C/min in the oven. The humidity chamber can only regulate with accuracy at a constant humidity and temperature; therefore, the modulus was determined at 5 °C temperature steps from 25 to 60 °C and at 10% relative humidity steps from 10 to 80%RH.
2.5 Water content measurements in immersed fibres
2.6 Tensile behaviour
Quasi-static tensile tests were carried out on yarn samples on a 10 kN capacity Instron™ test frame, at a loading rate of 50 mm/min. Strain was measured using two digital cameras to measure the displacements of two markers fixed to the sample, separated by a known distance (typically 300 mm).
Tensile creep tests were performed using the DMA equipment described above.
3 Nylon reference behaviour
To understand the relationship between water uptake and tensile behaviour, these aspects have to be studied separately first. Therefore, this section is devoted to the definition of glass transition temperature, water sorption and mechanical behaviour of dry fibres.
3.1 Glass transition temperature
The results obtained show a typical polyamide response. The glass transition temperature value was determined by considering the drop in storage modulus E′. In the reference state (0%RH) the Tg value reaches 55 °C on average, which is a little higher than for bulk polyamide 6 (around 50 °C).
3.2 Water sorption
To establish water sorption kinetics in fibres surrounded by a humid environment rather than liquid water the DVS technique is used. This testing method records the weight variation with a constant time step during tests where the relative humidity is controlled and progressively increased by steps. These tests were performed on short lengths of yarns (5 mm) of nylon 6.
To establish these parameters, we use a two-step identification. First, we consider that (H × a) is predominant compared to the power term and we identify H0 and Ea from the curve of ln(H) vs. 1/T. Then, we subtract the linear term from the initial maximal moisture content and identify b and m over the second part of the curve. These parameters appear to be independent of the temperature as previously shown in Arhant et al. (2016) and Broudin et al. (2015). In the case of nylon 6 fibres we obtain the following parameters: H0 = 1.39; Ea = − 7 kJ/mol1; b = 3; m = 7. If we refer to the published values on bulk polyamide 6 materials in Arhant et al. (2016) we can note that the values of activation energy (Ea) and b are very similar to the published parameters. Nevertheless, in the case of PA6 fibre materials the pre-exponential factor (H0) and m both differ from the parameters of bulk specimens. The polymer chains in fibre materials are usually more oriented and their crystallinity ratio is higher, this affects the maximal water uptake (m and H0) significantly.
Contrary to the case of water diffusion in a humid environment, when the samples are immersed there is no standard device that can record the weight variation with the same accuracy as DVS. Therefore, the DSC method described in Sect. 2 above was used to identify the different water quantities: inside filaments, between fibres and in total.
The samples were pre immersed in deionised water at 25 °C for different times and then analyzed by DSC.
In the case of immersion we cannot describe the transient part of the diffusion curve since the exposure time would be too close to the diffusion characteristic time. Therefore, the handling time becomes critical, and has an impact on the water uptake values obtained within the filaments.
Finally, the maximum moisture content in immersed filaments (10%) is compared with the values from humid environment exposure and the model results shown previously for a humid environment, Fig. 5. This comparison highlights a clear difference between the diffusion behaviour immersed and in humid environment. The model established for relative humidity diffusion cannot therefore accurately predict the value in the case of immersion. It would be interesting to perform additional tests in humid environments closer to 100%RH, in order to compare these experimental values both to Henry’s law and to the results for immersed fibres.
3.3 Mechanical behavior
3.3.1 Quasi-static tension
From 0 to 3.5% strain, the curve reflects an alignment of the amorphous regions, the crystalline domain remaining unloaded. As a consequence, amorphous regions are partly decreased, by the crystallisation induced by mechanical stress, the crystallinity ratio increases significantly (Marcellan et al. 2006).
From 3.5 to 10% strain, the characteristic shape of the end of the amorphous domain orientation appears. In this part, the crystallinity ratio increases more slowly (Marcellan et al. 2006).
From 10 to 15% strain, enthalpy mechanisms are activated: C–C bonds form crystals and oriented regions are elongated. The chains from crystalline regions start to be affected by the loading and then break.
Above 15% strain, the visco-plastic deformation mechanism is related to macro-fibrils sliding.
These micro and macro-damage mechanisms result in a ductile failure of the polyamide fibre.
3.3.2 Tensile creep
In service, mechanical loads are rarely quasi-static. For example, a mooring line may be subjected to a constant mean load with cyclic amplitude variations due to waves. In this section the creep behaviour of these fibres will be characterized, as a first step towards investigating their long term response.
Nylon polymer shows a logarithmic viscoelastic/viscoplastic creep behaviour (Hunt and Darlington 1978, 1979, 1980). To model this type of mechanical response under creep loading a thermodynamic model is used, based on Schapery’s theory (Schapery 1969).
In this set of equations σc is the stress applied on the specimen from 0 to t1 seconds and the parameters g0D0, g2D1, m, Dp are the identified viscoelastic/viscoplastic parameters. They respectively correspond to the instantaneous compliance (g0D0), the creep rate (g2D1), the instantaneous plasticity (m) and the plastic rate (Dp).
After identification of these parameters (detailed in Chailleux and Davies (2003, 2005) at different loading levels, interpolations were made on each parameter in order to predict their value for any applied stress (with polynomial functions).
In the study, we performed seven creep/recovery cycles from 10 to 70 N, which corresponds to a range from 58 to 409 MPa (Fig. 8). These stresses correspond to the usual range of application of nylon ropes (below 50% of their ultimate stress).
An example of the results obtained for nylon 6 at 25 °C and 0%RH is shown in Fig. 8. This gives information on the creep/recovery behaviour of dry nylon 6. It is clear that there is a significant residual deformation after unloading.
In practice, most applications induce some relative humidity or immersion in water, therefore the results from the same identifications on nylon in humid environments are detailed in the next section.
4 Influence of water on nylon 6 fibres
First, the influence of water on the glass transition behaviour will be discussed, then effects on mechanical response in tension and creep will be presented.
4.1 Glass transition temperature
The glass transition temperature is a physical parameter that is very sensitive to water uptake.
4.1.1 Measurements and identification
The range of temperatures that can be studied is smaller than in the 0%RH case; tests were performed from 25 to 50 °C.
4.1.2 T g prediction
To summarise, combining the equations Eqs. (7)–(9) provides the relationship between the glass transition temperature of the polymer (Tg) and water uptake in the polymer (C(t)). Moreover, in Sect. 3.2 we related the moisture water uptake in the polymer to the water activity (a) (Eq. 5). Therefore, by replacing C(t) in Eq. (8) by C∞(a) for each temperature and relative humidity, we can predict the variation of Tg with relative humidity according to Simha–Boyer theory (red line).
To conclude, the Simha–Boyer prediction (red line) and experimental data (black points) in Fig. 11 show a good correlation. This observation is true even for the highest relative humidity values (80%RH).
4.2 Hygroscopic expansion
The other major physical change induced by water in the polymer is hygroscopic expansion. This phenomenon is created by the insertion of the water molecules between the polymer chains, which generates swelling.
Some previous results are available on this subject for bulk (Adamson 1980) and fibre (Kawasaki et al. 1962) polymers. A reasonably linear dependence of hygroscopic elongation on the water uptake was observed, with a threshold value. For the water content close to the saturation value, this relation does not appear to be verified, probably due to clustering effects in the polyamide (Kawasaki et al. 1962); for high water content the molecules start to gather in clusters and therefore the increase in water uptake will generate lower hygroscopic elongation than in the case of isolated molecules.
The results of the present study highlight the same trend, with a hygroscopic coefficient (β) equal to 0.488. This linear law is verified up to a 5% water uptake in the fibres, nevertheless we do not have the threshold value since it appears to be reached for relative humidity lower than 20%RH.
Further studies should aim to determine the threshold value for polyamide 6, but this type of test will require the development of a special test device, which can maintain low relative humidities in the range from 0 to 20%RH.
4.3 Quasi-static tensile behaviour
The stiffness and break strength are reduced by water diffusion, which is due to reversible plasticization effects. Nevertheless, these changes are still small compared to results for bulk polyamide in the literature (Silva et al. 2013; Launay et al. 2013; Miri et al. 2009; Boukal 1967, for which very large changes have been noted. Short yarn samples are saturated with water after 15 min of exposure. However, the time required for sample preparation and performing tensile tests on yarns is approximately 5 min, and therefore the fibres will start to desorb significantly before testing. This suggests that the effects of water on the quasi-static mechanical response obtained for the aged yarns here may be underestimated.
4.4 Tensile creep behaviour
The creep strain results (Fig. 14) obtained before identification show the large increase in the initial value and creep rate induced by water uptake. The recovery data show a permanent elongation increasing with the relative humidity. These results are not unexpected; as we observed previously, water uptake induces significant plasticization (reversible decrease in Tg) and therefore makes the material more ductile (for the conditions of this study). However, this increase in strain may have a large effect on the station-keeping ability of mooring lines as it will affect the apparent stiffness, and must be considered in rope design.
4.5 Identification of model parameters
In order to describe the influence of humidity more accurately, and to set up a framework to predict long term strain, a viscoplastic/viscoelastic model has to be identified. In this work we consider a modified Schapery model to define this viscous behaviour (see part 3). For each condition (stress and humidity) the different parameters of the law have been identified, following the same method as for the dry fibres in part 3.
These plots show that the model parameters, elastic, viscoelastic and viscoplastic, are all affected by both applied stress and moisture in a complex manner. This indicates that in order to predict the response of a mooring line it will be necessary to perform a complete characterization. These variations in long term strain will affect the effective line stiffness, and will need to be included in any analysis. They are specific to nylon fibre ropes; most other fibres used for mooring lines are much less sensitive to moisture. The consequences for the line may be positive if they allow larger displacements to be absorbed, or negative if stiffness proves insufficient to keep the device on station.
Further tests are underway in order to develop and validate this modelling approach and to extend it to both fully immersed polyamide yarns and to rope samples. The latter undergo an additional constructional stretch due to the rope geometry, which must also be analysed and included in the model.
Nylon fibres are particularly sensitive to moisture, which reduces the Tg of the amorphous structure. This can be predicted, and has a small influence on the static tensile properties but a significant effect on the long term behaviour. It is therefore essential to quantify the influence of moisture on the creep response of nylon 6 fibres if they are to be used in long term marine applications such as mooring of floating wind turbines. A simple model summing elastic, viscoelastic and viscoplastic strain components, could prove useful in quantifying these effects.
Funding was provided by Ifremer and Université de Nantes (Internal funding).
Compliance with ethical standards
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
- Ridge IML, Banfield SJ, Mackay J (2010) Nylon fibre rope moorings for wave energy converters. In: Proceedings MTS/IEEE OCEANS 2010 conference, Seattle, 2010, IEEE, New Jersey, US, pp 1–10Google Scholar