Accelerated aging test of hydrogenated nitrile butadiene rubber using the time–temperature–strain superposition principle

Abstract

The effect of physical aging on the long-term performance of hydrogenated nitrile butadiene rubber (HNBR) is very important to its engineering applications. In this paper, aging tests of HNBR have been conducted systematically at different temperatures and levels of uniaxial tensile strain. Fracture strain was chosen as the index of the severity of aging. Applied strain shows an equivalent effect on aging acceleration to that of temperature; the higher the temperature and strain level are, the faster the aging is. A WLF-type time–temperature–strain superposition principle (TTSSP) is proposed to study the long-term aging performance of HNBR with short-term test data. With the constructed master curves, time and cost can be significantly reduced for the evaluation of the long-term aging behavior of the rubber materials.

---

1. Introduction

Due to its low weight, significant hyperelasticity and good damping characteristics, hydrogenated nitrile butadiene rubber (HNBR) has been utilized for a variety of engineering applications such as in shock absorbers, seals and hoses in pipelines and in the automotive industry. During the service period of these rubber components, even without considering chemical degradation, aging does occur and can be accelerated by severe service conditions such as elevated temperature, applied mechanical loading etc. The physical aging process directly leads to the degradation of various mechanical properties of rubber such as its tensile,^1–3 compression^4,5 and even tribological properties.⁶ Meanwhile, it also influences the long-term properties, for example, by accelerating the stress relaxation rate of seal materials.^7,8

As it is widely utilized, HNBR’s service life prediction and evaluation are very important for both design and service procedures to ensure the safety and reliability of the rubber components. The degradation and aging mechanisms of HNBR have attracted a great deal of attention.^9–12 Since the service lifetime of rubber components can be as long as many years, a reliable accelerated test method is necessary. Much work has been done to study the aging behaviour as well as the aging mechanism of various rubbers at elevated temperature.^13–17 It has been generally accepted that temperature can accelerate the aging process by expediting the reaction rate and promoting molecular mobility. Using an elevated temperature to substitute for a long period of time,^18–22 the WLF equation is widely adopted for the time–temperature superposition principle (TTSP) to predict long-term aging behaviour.^23,24

Meanwhile, the accelerating effect of applied mechanical loading on aging has also been observed. Xiong et al. found that the aging process of nitrile butadiene rubber was significantly accelerated under different types of loading conditions.²⁵ Gui et al. described the accelerated aging of poly(methyl methacrylate) under tension.²⁶ Budrugeac demonstrated that the thermal aging of nitrile butadiene rubber was expedited tremendously under air pressure.²⁷ Zhu et al. discussed the possible mechanisms of accelerated aging of polymers under applied stress.²⁸ A time–temperature–stress superposition principle,²⁹ borrowing the idea from accelerating studies of long-term creep,^30–37 has been successfully proposed to study the acceleration effect of applied stress on the aging of polycarbonate. While little work can be found in the literature concerning the effect of applied strain on the aging process, many rubber components are actually subjected to a prescribed strain rather than stress for certain applications, for example as seals and gaskets. Also, for the purposes of experimental operation, strain can be more conveniently/precisely controlled than stress due to rubber’s hyperelasticity.

In this paper, to understand the long-term aging behaviour of HNBR under applied strain, a WLF-type time–temperature–strain superposition principle (TTSSP) was proposed. The aging tests of HNBR were systematically conducted at different temperatures and uniaxial tensile strain levels. After constructing the master curves of HNBR aging, the long-term aging behaviour can be predicted by using the short-term test data at elevated temperature and strain level. Dramatically saving time and cost, applied strain could be a good alternative accelerating approach to assess the long-term aging behaviour of rubber materials.

2. Theoretical foundation

It is well known that the aging and degradation of rubber is a continuous process and is strongly temperature dependent. TTSP can be employed to describe the relationship between time t and temperature T.^38,39 The widely adopted WLF-type equation is applicable in the temperature range T > T_g. The aging rate is assumed to be linearly dependent on the temperature change:

τ = τ₀ + α_T(T − T₀) (1)

where τ₀ is the aging rate at the reference temperature T₀, and α_T is the coefficient of the temperature-induced accelerated aging rate. Following a similar approach to previous reports,^23,24 the time–temperature shift factor of the WLF-type TTSP is:where C₁ and C₂ are material constants and their values can be chosen in the same manner as in the literature.^40,41

It is a generally accepted fact that the higher the aging temperature is, the higher the molecular chain mobility is. The enhanced mobility of the molecular chains leads to the HNBR reaching an equilibrium state of molecular structure more rapidly, i.e., aging faster than is the case at low temperature. Concerning the mobility of the molecular chains, applied strain similarly accelerates the aging process by lowering the energy barrier of molecular chain movement. Meanwhile, in many practical engineering applications strain is applied to rubber components during their service lifetime, so strain is not only a good alternative factor but is also necessary for accelerated aging tests.

Assuming the aging rate is also linearly dependent on the applied strain level, the aging rate considering the temperature and strain simultaneously can be formulated as:

τ = τ₀ + α_T(T − T₀) + α_ε(ε − ε₀) (3)

where α_ε is the coefficient of the strain-induced accelerated aging rate. The combined shift factor α_Tε for the time–temperature–strain superposition principle (TTSSP) iswhere C₃ is a material constant and its value can be chosen in the same manner as in the literature.³⁷

For the unstrained cases, the proposed TTSSP reverts to the traditional time–temperature superposition principle. Meanwhile, for a constant aging temperature, i.e., T = T₀, the time–strain superposition principle (TSSP) can be used, and the time–strain shift factor α_ε can be written as:

Due to its inherent nature, the material properties of rubber change with temperature as well as the applied strain level.^43,44 A vertical shift factor is necessary to construct the aging master curve:

log b_Tε = D(T − T₀) + log ε (6)

where D is a material constant with the dimension of 1/°C and is taken as 0.025 in this paper. Thus, the aging test of the HNBR can be carried out in a time-efficient and cost-effective fashion with the help of the proposed TTSSP approach.

3. Experimental

3.1. Material and specimens

90% saturated HNBR containing 50% acrylonitrile was supplied by Kaidi Northwest Rubber Co. Ltd in sheet form with a thickness of 2 mm. This was punched into dumb-bell specimens according to ISO 37-2011. The specimen size is shown in Fig. 1.

Fig. 1 The specimen size for the aging test (unit: mm).

3.2. Experimental setup

The aging tests were conducted in an environmental chamber (WG03, INBORN Experiment Instrument CO., LTD, Chongqing). A home-made loading apparatus was utilized to apply the designated level of tensile strain to the specimens. The physical aging conditions are listed in Table 1. While a 30% applied strain is already capable of inducing a significant accelerated aging effect, in this work we chose 60% as the maximum applied strain level for the convenience of conducting the experiment. The temperature range was chosen as 60 °C to 100 °C, in accordance with previous similar works on the thermal aging of HNBR.^11,25 At least three specimens were tested under each aging condition. After reaching the prescribed aging time, the specimens were stored in air for more than 16 hours. Then the length was recorded as the initial length of the specimen before the tensile test.

Table 1 Aging conditions

Aging temperature (°C)	Applied strain (%)	Aging time (days)
60	0	1, 3, 6
	30	1, 2, 3, 5
80	0	1, 2, 4, 6
	30	1, 2, 3, 5
	60	1, 2, 3, 5
100	0	1, 2, 4, 7
	30	1, 2, 4, 6
	60	1, 2, 3, 4, 5

---

Uniaxial tensile tests were performed to evaluate the change in the mechanical properties of HNBR after aging. All the tensile tests were conducted at room temperature (about 25 °C) using an electronic universal testing machine (WDW3100, Changchun Kexin Testing Instrument Co., Ltd, Changchun) at a 500 mm min⁻¹ loading rate following the requirement of ISO 37-2011 until the specimen was broken. Fracture strain, also known as elongation at break, or the ratio between the elongation recorded at the breakage and the initial length, was chosen as the index to describe the severity of the physical aging in this paper.

4. Results and discussion

4.1 Effect of temperature on HNBR aging with constant strain level

Fig. 2 and 3 show the effect of temperature on HNBR aging at two prescribed strain levels, i.e., 0% and 30%, respectively. Three different aging temperatures, 60 °C, 80 °C and 100 °C were used. It should be noted that the fracture strain of unaged (0 day) virgin samples, which is around 480% for the studied HNBR, is not given in the figures because of the logarithmic time coordinate.

Fig. 2 Aging behaviour of HNBR at different temperatures under a strain level of 0%: (a) fracture strain versus aging time; (b) master curve at a reference temperature of 60 °C.

Fig. 3 Aging behaviour of HNBR at different temperatures under a strain level of 30%: (a) fracture strain versus aging time; (b) master curve at a reference temperature of 60 °C.

It could be found from Fig. 2(a) and 3(a) that the fracture strain of HNBR decreases with the progression of the aging process at different aging temperatures. This means that the mechanical properties of rubber do decay with storage time. Meanwhile, the fracture strain at higher temperature is smaller than that at lower temperature with the same aging time, either with or without applied strain. In both the evolution trends and the actual values of the fracture strain after aging, this paper shows a good consistency with Xiong’s work.²⁵

Using the TTSP approach, i.e., eqn (2), as well as eqn (6), the aging master curves were constructed. The material parameters used for the time–temperature shift, vertical shift and time–strain shift are listed in Table 2. Fig. 2(b) and 3(b) illustrate the evolution of the fracture strain of HNBR under two strain levels (0% and 30%) at a reference temperature of 60 °C using the short-term test data at higher temperature. For the unstrained case, the aging behaviour of HNBR for about 32 days is predicted using the short-term (7 days) data. As for the applied strain of 30%, the tendency of decreasing fracture strain with aging time for around 28 days is obtained using the aging data of 6 days.

Table 2 Material parameters for the time–temperature–strain superposition principle

Parameters	60 °C	80 °C	100 °C
C1	4.2	4.2	4.2
C2	210	210	210
C3	2.25	1.15	0.792
D	0.025	0.025	0.025

---

4.2 Effect of strain on HNBR aging with constant temperature

Fig. 4 and 5 show the accelerated aging behaviour of HNBR at three strain levels (0%, 30%, and 60%) and two constant aging temperatures: 80 °C and 100 °C, respectively. It is obvious from Fig. 4(a) and 5(a) that the fracture strain decreases with aging time for both aging temperatures. The acceleration effect of applied strain is as significant as, if not more remarkable than, that of temperature in the studied ranges of temperature and strain level. In addition, the higher the applied strain level is, the faster the fracture strain drops and the more severe the aging of HNBR is. In other words, HNBR ages much faster when subjected to an applied strain. This finding also agrees with the experimental results from Xiong.²⁵

Fig. 4 Aging behaviour of HNBR under different strain levels at 80 °C: (a) fracture strain versus aging time; (b) master curve at a reference strain level of 0%.

Fig. 5 Aging behaviour of HNBR under different strain levels at 100 °C: (a) fracture strain versus aging time; (b) master curve at a reference strain level of 0%.

Then, the time–strain superposition principle (TSSP) was adopted to construct the master curves of the strain accelerated physical aging of HNBR with the help of eqn (5) and (6). As shown in Fig. 4(b) and 5(b), master curves at a reference strain level of 0% have been established for 80 °C and 100 °C, respectively. For a temperature of 80 °C, the TSSP can predict the aging behaviour of unstrained HNBR for up to 130 days using the short-term (6 days) data with applied strain. As for the aging temperature of 100 °C, the decrease of fracture strain with an aging time of almost one year can also be predicted.

It should be noted that although this work only investigates the tensile properties of HNBR, the degradation of the material properties after the accelerated aging process shows the same tendency as in literature reports in which the compression properties declined dramatically after aging.^4,5

4.3 Time–temperature–strain superposition on HNBR aging

From the above results, elevated temperature and applied strain clearly show equivalent acceleration effects on the physical aging process of HNBR. Both factors can substitute for a long aging time. To consider the effects of time, temperature and strain on the long-term aging behaviour of HNBR simultaneously, the proposed time–temperature–strain superposition principle (TTSSP) can be employed to efficiently construct a master curve under arbitrary reference conditions with the help of eqn (4) and (6).

The test results of all eight aging conditions are presented in Fig. 6(a). The long-term degradation of fracture strain under the reference conditions of 60 °C with zero strain is shown in Fig. 6(b). The long-term aging behaviour of HNBR for up to 550 days can be predicted using the short-term (<10 days) test results.

Fig. 6 Aging behaviour of HNBR under various aging conditions: (a) fracture strain versus aging time; (b) master curve under reference conditions of 60 °C with 0% strain.

When applying elevated temperature and strain simultaneously, an accelerated aging effect which is more tremendous than that due to only temperature or strain,^5,26 can be obtained. Thus substantial time and cost can be saved for the study of the long-term aging behaviour of HNBR.

Although stress relaxation under constant strain^42–47 does exist and neglecting it may lead to an over-estimation of aging, the tendency of the aging and degradation of rubber has been satisfactorily predicted using the proposed TTSSP approach for the chosen range of temperature and strain. The possible influence of relaxation on accelerated aging in severer aging conditions should be investigated in future work.

5. Summary

Systematic aging tests of HNBR have been performed to study the acceleration effect of temperature and strain level. A time–temperature–strain superposition principle was proposed to evaluate the long-term aging performance of HNBR using short-term test results.

(1) Applied strain shows an equivalent accelerating effect on the aging of HNBR to that of elevated temperature.

(2) Using fracture strain as the index of the material’s performance, a master curve was constructed successfully to evaluate the tendency of the long-term aging process.

(3) With the proposed WLF-type time–temperature–strain superposition principle, time and cost can be substantially saved by evaluating the long-term aging behaviour of rubber with short-term test data.

Acknowledgements

This work is supported by National Natural Science Foundation of China (11172249, 11472231). The authors also would like to thank the partial financial support from Ministry of Education of China (NCET-12-0938), Science and Technology Department of Sichuan Province (2013JQ0010) and State Key Laboratory of Polymer Materials Engineering of China (Sichuan University) (KF201204).

Notes and references

1. H. Wei, L. Guo, J. Zheng, G. Huang and G. Li, RSC Adv., 2015, 5, 62788–62796.
2. J. Feng, Q. Zhang, Z. Tu, W. Tu, Z. Wan, M. Pan and H. Zhang, Polym. Degrad. Stab., 2014, 109, 122–128.
3. P.-Y. le Gac, V. le Saux, M. Paris and Y. Marco, Polym. Degrad. Stab., 2012, 97, 288–296.
4. F. Guo, X. Jia, L. Huang, R. F. Salant and Y. Wang, Polym. Degrad. Stab., 2013, 98, 2193–2200.
5. P. Morrell, M. Patel and A. Skinner, Polym. Test., 2003, 22, 651–656.
6. C. Dong, C. Yuan, X. Bai, X. Yan and Z. Peng, Wear, 2015, 322, 226–237.
7. J. Zhao, R. Yang, R. Iervolino, B. van der Vorst and S. Barbera, Polym. Degrad. Stab., 2015, 115, 32–37.
8. K. Xiang, S. Wu, G. Huang, J. Zheng, J. Huang and G. Li, Macromol. Res., 2014, 22, 820–825.
9. S. Bhattacharjee, A. K. Bhowmick and B. N. Avasthi, Polym. Degrad. Stab., 1991, 31, 71–87.
10. E. Campomizzi, H. Bender and W. von Hellens, Kautsch. Gummi Kunstst., 2001, 54, 114–121.
11. R. J. Pazur, J. Cormier and K. Korhan-Taymaz, Rubber Chem. Technol., 2014, 87, 53–69.
12. D. Yang, M. Tian, Y. Dong, H. Liu, Y. Yu and L. Zhang, Smart. Mater. Struct., 2012, 21, 035017.
13. M. B. Hassine, M. Naït-Abdelaziz, F. Zaïri, X. Colin, C. Tourcher and G. Marque, Mech. Mater., 2014, 79, 15–24.
14. K. Reincke, B. Langer, S. Doehler, U. Heuert and W. Grellmann, Kautsch. Gummi Kunstst., 2014, 67, 60–67.
15. T. Nakazono and A. Matsumoto, J. Appl. Polym. Sci., 2010, 118, 2314–2320.
16. Z. X. Ooi, H. Ismail and A. A. Bakar, J. Appl. Polym. Sci., 2013, 130, 4474–4481.
17. M. Santoso, U. Giese and R. Schuster, KGK, Kautsch. Gummi Kunstst., 2007, 60, 192–198.
18. K. T. Gillen, R. Bernstein and M. H. Wilson, Polym. Degrad. Stab., 2005, 87, 257–270.
19. K. Alwis and C. Burgoyne, Appl. Compos. Mater., 2006, 13, 249–264.
20. L. C. Bank, T. R. Gentry and A. Barkatt, J. Reinf. Plast. Compos., 1995, 14, 559–587.
21. E. J. Barbero and K. J. Ford, J. Eng. Mater. Technol., 2004, 126, 413–419.
22. D. Hukins, A. Mahomed and S. Kukureka, J. Biomed. Eng., 2008, 30, 1270–1274.
23. J. D. Ferry, Viscoelastic properties of polymers, John Wiley & Sons, 1980.
24. M. L. Williams, R. F. Landel and J. D. Ferry, J. Am. Chem. Soc., 1955, 77, 3701–3707.
25. Y. Xiong, B. Q. Fu, S. Y. Guo and Z. Lu, Equip. Environ. Eng., 2012, 9, 52–55.
26. S. Z. Gui and Y. Nanzai, Polym. J., 2001, 33, 444–449.
27. P. Budrugeac, Polym. Degrad. Stab., 1995, 47, 129–132.
28. Z. P. Zhu, H. Pi, S. Y. Guo and G. X. Li, Polyvinyl Chloride, 2008, 36, 6–9.
29. C. Jiang, H. Jiang, Z. Zhu, J. Zhang, S. Guo and Y. Xiong, Polym. Eng. Sci., 2015, 55, 2215–2221.
30. A. E. Akinay, W. Brostow, V. M. Castaño, R. Maksimov and P. Olszynski, Polymer, 2002, 43, 3593–3600.
31. F. C. Chang, F. Lam and J. F. Kadla, Mech. Time-Depend. Mater., 2013, 17, 427–437.
32. S. Jazouli, W. Luo, F. Bremand and T. Vu-Khanh, Polym. Test., 2005, 24, 463–467.
33. G. H. Kim and H. Lu, Polym. Test., 2007, 26, 839–845.
34. G. H. Kim and H. Lu, Polym. Test., 2008, 27, 114–121.
35. J. Zhang, H. Jiang, C. Jiang, G. Kang and F. Lu, Polym. Test., 2015, 44, 8–14.
36. R. G. Jin and W. Liu, J. Beijing Univ. Chem. Technol., 1994, 2, 18–25.
37. W. B. Luo, T. Q. Yang and Q. L. An, Acta Mech. Solida Sin., 2001, 14, 219–224.
38. R. Andrews and A. Tobolsky, J. Polym. Sci., 1951, 7, 221–242.
39. P. C. Hiemenz and T. P. Lodge, Polymer chemistry, CRC press, 2007.
40. A. K. Doolittle, J. Appl. Phys., 1951, 22, 1471–1475.
41. Y. Miyano, M. Nakada and R. Muki, Mech. Time-Depend. Mater., 1997, 1, 143–159.
42. I. M. Ward and J. Sweeney, Mechanical properties of solid polymers, John Wiley & Sons, 2012.
43. S. Clarke, F. Elias and E. Terentjev, Eur. Phys. J. E, 2000, 2, 335–341.
44. R. Kaltseis, C. Keplinger, S. J. A. Koh, R. Baumgartner, Y. F. Goh, W. H. Ng, A. Kogler, A. Tröls, C. C. Foo and Z. Suo, RSC Adv., 2014, 4, 27905–27913.
45. X. Wang, Y. Wu, Q. Li, T. W. Chan, L. Zhang and S. Wu, RSC Adv., 2015, 5, 66168–66177.
46. A. S. Khan, M. Baig, S. Hamid and H. Zhang, Int. J. Solids Struct., 2010, 47, 2653–2659.
47. T. Tada, K. Urayama, T. Mabuchi, K. Muraoka and T. Takigawa, J. Polym. Sci., Part B: Polym. Phys., 2010, 48, 1380–1387.

---