Elsevier

Polymer Testing

Volume 78, September 2019, 105979
Polymer Testing

Test Method
Comparison of short and long term creep testing in high performance polymers

https://doi.org/10.1016/j.polymertesting.2019.105979Get rights and content

Highlights

  • Accelerated creep tests can be used to compare the relative creep response of different materials at elevated temperatures.

  • We predicted the creep behaviour for amorphous polyetherimide with a Tg of 217 °C, at temperatures of 120 °C and 150 °C.

  • We produced an approximation for the creep response of PEEK with a Tg of 143 °C, both above and below Tg.

  • The stepped isostress method is a useful technique comparing polymers of complex and simple rheology at high temperatures.

  • For both semi-crystalline and amorphous plastics, we predicted the creep response to over 70 times the test time.

  • Amorphous PAI with a Tg of 275 °C has very high creep resistance at 120 °C and 150 °C, compared to PEEK and PEI.

Abstract

The creep behaviour of thermoplastics plays an important role in engineering design and especially the failure and lifetime prediction of structural components in a variety of industries. Often this mechanical property is determined with long-term creep tests. In order to reduce the development time of components made from such materials used in applications with conditions over 100 °C for long times, we propose the use of dynamic-mechanical analysis to predict long-term creep deformation through the stepped isothermal (SIM) and the stepped isostress methods (SSM). The present study aims to determine whether SIM and SSM can be generalised to material characterisation of advanced polymers at high temperature. Both methods are evaluated and verified by long-term measurements. Both amorphous and semi-crystalline polymers are studied and the influence of the glass transition temperature (Tg) on the accelerated test results is examined with tests conducted above and below the Tg region for PEEK. The results of our analysis show accelerated testing is an effective tool in predicting long term behaviour and comparing the creep response of thermoplastics. PAI exhibits a far superior creep response when compared to PEEK and PEI over 100 °C.

Introduction

Polymer compounds are increasingly used in structural applications, which require strict dimensional stability, minimal deformation with time and strong temperature resistance. They are inherently viscoelastic materials with time dependent mechanical properties and understanding the creep behaviour is important for various load bearing applications, particularly in the early stages of component development [1]. Long-term data on mechanical properties are therefore required to understand the material deformation over time for application in the aerospace, automotive and construction industries. Uncertainty in the long-term properties often results in exaggerated safety factors and hence more material usage than necessary or less suitable material alternatives being preferred. Long-term tests used to measure creep behaviour are expensive and by nature, require long times to carry out. The time-temperature and time-stress superposition principles underpin accelerated characterisation procedures, which allow the long-term creep behaviour to be predicted in a fraction of the time. This reduces the costs associated with long product development cycles and their subsequent impact on the supply chain through reducing product lead times [2]. We propose to carry out accelerated creep experiments using Dynamic Mechanical Analysis, which has been effectively demonstrated in past work [3].

Increasing the temperature of a material is method commonly used to accelerate the creep rate of polymers and hence it is often used to characterise long-term creep behaviour in shorter times. The time-temperature superposition principle is based on the Doolittle formula [4], which gives the viscosity of a liquid as a function of free volume. Williams, Landel and Ferry [5] extended this dependency to relaxation behaviour in amorphous polymers and glass forming liquids above the glass transition temperature, through the following equation:log(aT)=C1(TT0)C2+(TT0)where C1 and C2 are material constants, aT is the temperature shift factor and T0 is the reference temperature. For conditions below the glass transition temperature, an Arrhenius relationship is commonly used to define the dependency of the shift factor on temperature [6,7]:log(aT)=ΔH2.303R(1T1T0)where ΔH is the activation energy and R is the universal gas constant.

Free volume theory serves as the basis for time-temperature superposition, as the increase in the free volume allows tightly packed polymer chains to move past each other more freely, thus resulting in an apparent increase in viscoelasticity. This effect allows the observer to simulate the increases in free volume caused by deformation and thus accelerate creep behaviour. Short-term creep experiments at different temperatures are collated and used to generate a master curve through application of the temperature shift factor, which allows independent creep experiments to be translated in time through the relationship between time and temperature, using equations of the form (1-1) and (1-2). The time-temperature superposition principle is considered applicable to materials with simple rheology, in order that the Boltzmann superposition principle remains valid. Through modelling of molecular scale interactions in polymer materials, Brostow et al. [8] showed that at extremely small strains, near constant free volume exists in deformation, but that changes in deformation rates may exist between adjacent second phase agglomerates. However, it has been suggested that time-temperature superposition may be applicable to multiphase materials, provided no changes in the degree of crystallinity take place during testing and equations of state are used in modelling [3,9]. Brostow argued in Ref. [10] that the time-temperature superposition principle may be applicable to such materials, if an equation of the following form is applied in conjunction with the Hartmann equation of state:ln(aT)=A0+A1(v˜1)where A0 and A1are material constants and v˜ is the reduced volume. The time-temperature superposition principle was subsequently validated for multiphase PET/0.6PHB PLC using this concept. According to the Adam-Gibbs theory, the conformational entropy induced by segment motion disappears, at temperatures between 50K and 60K below Tg[11,12]. The change in creep mechanism at high temperature resulting from this effect may limit the applicability of the results at the highest temperature to the final master curve, which will be discussed in relation to our results in section 3.2. Time-temperature superposition has been used effectively in past research with commodity polymers under ambient conditions by several authors, with a comprehensive review carried out in Ref. [13].

Thornton et al. [14] proposed the stepped isothermal method (SIM) to predict the long-term creep behaviour of geogrids through testing a single sample under constant load. The sample is placed under constant stress and the temperature is increased step wise to accelerate the creep rate using time-temperature superposition. Until recently, this method had been applied only to high modulus yarns and fibres, which with high surface area to volume ratios can be rapidly heated, without inducing temperature profiles and undefined creep conditions in the sample. Achereiner et al. [15] were successfully able to demonstrate the validity of SIM in thick polypropylene samples and introduced a graphical approach to generate master curves from tests on a single sample.

The time-stress superposition effect was first observed experimentally by O'Shaughnessy [16] as an analogue to the effect produced by temperature, in increasing the creep rate of viscoelastic materials. The free volume interpretation of the temperature effects on glassy and semi-crystalline materials was found by Ferry and Stratton [17] to extend to pressure, concentration and tensile stress. This observation forms the basis of the time-stress superposition principle because it demonstrates the dependency of free volume on applied stress in polymers, in much the same way as temperature. Lai and Bakker [18] used time-stress superposition to successfully predict the highly non-linear creep behaviour of HDPE, suggesting that non-linear characterisation of semi-crystalline materials is possible. Wang et al. [19] were able to prove this similarity mathematically and showed the stress shift factor can be related to the accelerating stress using a relationship of the same form as the Williams-Landel-Ferry equation:log(aσ)=C1(σσ0)C2+(σσ0)

The mathematical equivalence builds on the work of Brostow in Ref. [20], who used the Hartmann equation of state and free volume based parameters to represent mechanical and rheological behaviours in an analogue of the temperature equivalence effect and suggested the following equation:ln(aσ)=ln(v(σ)vref)+B[(v˜1)1(v˜ref1)1]+C(σσref)

Brostow subsequently validated equations (1)–(5) for the PET/0.6PHB PLC material in Ref. [21] at 20 °C and showed the stress shift factor could be determined from small data samples [22].

The use of stress to accelerate creep behaviour has some distinct advantages over the use of temperature. It allows the sample to remain at constant temperature, which reduces the likelihood of temperature induced chemical changes occurring in the sample such as those discussed earlier, incurred by changes in conformational entropy. For semi-crystalline polymers which exhibit structural changes with temperature, the physical and mechanical properties vary, particularly at the interface between phases where there exists uncertainty as to the exact mechanism which occurs during deformation at high temperature. The time-stress superposition principle is expected to produce more accurate predictions of long-term creep behaviour at high temperatures, assuming the degree of crystallinity remains constant [23].

The stepped isostress method (SSM), was proposed by Burgoyne et al. [24] for the testing of high modulus aramid fibres, and was successfully used to predict the creep response of Kevlar-49. The SSM method is based on the stepped isothermal method but overcomes many of the limitations associated with using temperature to accelerate the creep rate. This method has been successfully demonstrated by several authors in predicting long-term creep performance at room temperature, with commodity and commercial grade polymers [[25], [26], [27]]. With the development of new high performance materials such as Torlon PAI, which is typically used in high temperature applications, it is important to establish an accurate method for predicting long term creep behaviour under such operating conditions.

It is the aim of this work to assess the validity of the stepped isothermal and stepped isostress methods, in accurately predicting the creep behaviour of high performance and ultra polymers at temperatures above 100 °C experimentally. Both amorphous PEI and PAI and semi-crystalline PEEK are studied and the influence of the Tg region on the accelerated test results is examined. Tests are conducted above and below the Tg region for PEEK. The results of high temperature long-term tests, in addition to literature data, are compared to the accelerated tests conducted using Dynamic Mechanical Analysis. The long-term performance of test specimens is subsequently compared under various conditions to demonstrate how such a technique may be useful in comparing the creep resistance at high temperature.

Section snippets

Materials

The creep behaviour of three high performance polymers is evaluated in this study: polyetheretherketone (PEEK. 450G, Tg = 143 °C) from Victrex (Southampton, UK), polyetherimide (PEI, Ultem 1000, Tg = 217 °C) from Sabic (Riyadh, Kingdom of Saudi Arabia) and polyamideimide (PAI, Torlon 4203L, Tg = 275 °C) from Solvay (Brussels, Belgium).

Dynamical-mechanical analysis (DMA)

All accelerated test measurements were carried out using a DMA Q800 from TA Instruments with a 3-point bending configuration and analysed using associated

Comparing test methods at high temperature

In order to compare the long-term creep performance at high temperature, the accelerated testing techniques were first assessed and validated against long-term data under the proposed conditions. Ultem 1000 polyetherimide was used as a reference specimen for amorphous plastics and Victrex PEEK 450G was used as a reference for semi-crystalline plastics. Due to the availability of literature data and the differing properties of the samples under test, different conditions were used for this study.

Conclusions

Accelerated creep tests are required for many different applications in which knowledge of the long-term mechanical properties is critical. Several past works have investigated the stepped isothermal and stepped isostress methods for such a purpose with much success. The present work has shown that accelerated creep tests can be used to compare the relative creep response of different materials at elevated temperatures. We were able to accurately predict the creep behaviour for amorphous

Acknowledgements

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

References (30)

  • M. Tajvidi et al.

    Time-temperature superposition principle applied to a kenaf-fiber/high- density polyethylene composite

    J. Appl. Polym. Sci.

    (2005)
  • A.K. Doolittle

    Studies in Newtonian flow. II. the dependence of the viscosity of liquids on free-space

    J. Appl. Phys.

    (1951)
  • M.L. Williams et al.

    The temperature dependence of relaxation mechanisms in amorphous polymers and other glass-forming liquids

    Phys. Rev.

    (1955)
  • P.A. O'Connell et al.

    Arrhenius-type temperature dependence of the segmental relaxation below Tg

    J. Chem. Phys.

    (1999)
  • J. Kolarik

    Tensile creep of thermoplastics: time-strain superposition of non-iso free-volume data

    J. Polym. Sci., Part B: Polym. Phys.

    (2003)
  • Cited by (14)

    • A fast and precise methodology of creep master curve construction for geosynthetics based on stepped isothermal method (SIM)

      2021, Geotextiles and Geomembranes
      Citation Excerpt :

      Yeo and Hsuan (2010) used long-term and short-term TTS and SIM to study the creep behaviors of PET and PE-HD materials. Besides, Simoes et al. (2006), Kongkitkul and Tatsuoka (2007), Alwis and Burgoyne (2008), Hsiehl et al. (2008), Lainé et al. (2019), Fairhurst et al. (2019), and Achereiner et al. (2013) all these researchers have studied the tensile creep of geosynthetics using the SIM tests. For the other methods of accelerated creep, Tanks et al. (2017) evaluated the applicability of the SSM to the creep behaviors of unidirectional carbon fiber reinforced polymer (CFRP) lamina.

    • Developments of PEEK (Polyetheretherketone) as a biomedical material: A focused review

      2021, European Polymer Journal
      Citation Excerpt :

      In that case incorporation of reinforcement becomes necessary as unfilled PEEK is not suitable for use. Although in biomedical applications, usually, the operating temperatures are well below the glass transition temperature [6–11]. Now on a broader picture by the late 1990s, PEEK emerged as the leading high-performance thermoplastic polymeric material for replacing metal in the fields of implant especially in orthopedics and trauma [12–14].

    View all citing articles on Scopus
    View full text