talc composites

Accepted Manuscript Effect of Weathering-Induced Degradation on the Fracture and Fatigue Characteristics of Injection-Molded Polypropylene/Talc Composites Jung-Wook Wee, Min-Seok Choi, Hong-Chul Hyun, Ji-Hoon Hwang, ByoungHo Choi PII: DOI: Reference:

S0142-1123(18)30307-4 https://doi.org/10.1016/j.ijfatigue.2018.07.022 JIJF 4776

To appear in:

International Journal of Fatigue

Received Date: Revised Date: Accepted Date:

22 February 2018 12 July 2018 14 July 2018

Please cite this article as: Wee, J-W., Choi, M-S., Hyun, H-C., Hwang, J-H., Choi, B-H., Effect of WeatheringInduced Degradation on the Fracture and Fatigue Characteristics of Injection-Molded Polypropylene/Talc Composites, International Journal of Fatigue (2018), doi: https://doi.org/10.1016/j.ijfatigue.2018.07.022

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Effect of Weathering-Induced Degradation on the Fracture and Fatigue Characteristics of Injection-Molded Polypropylene/Talc Composites Jung-Wook Weea, Min-Seok Choia, Hong-Chul Hyunb, Ji-Hoon Hwangb, Byoung-Ho Choia,*,[email protected] aSchool

of mechanical engineering, Korea University, Seoul, 136-701, Republic of Korea

bGlobal

Technical Center, Samsung Electronics, Gyeonggi-do, Republic of Korea

*

Corresponding author.at: School of Mechanical Engineering, College of Engineering, Korea

University, 5-ga Anam-dong, Sungbuk-ku, Seoul, Republic of Korea

Highlights

Effect of weathering on fracture characteristics are analyzed by EWF tests. Variation of S-N characteristics of weathering-induced PP-talc composites is analyzed. Methodology of predicting fatigue lifetime considering progressive weathering is proposed.

Abstract Weathering-induced degradation causes a premature failure of polymeric materials in outdoor applications. In this study, the polypropylene and talc composites were degraded by accelerated and outdoor weathering. By comparing the degradation degree through the Fourier Transform-Infrared (FT-IR) analysis, the accelerated weathering factor was constructed. To investigate the effect of weathering on the short- and long-term mechanical properties, the

tensile, essential work of fracture, and fatigue tests were performed for pre-degraded specimens under accelerated weathering. The results demonstrated that the weathering-induced degradation slightly increased the elastic modulus and tensile strength; whereas such degradation dramatically reduced the strain at the break. The fracture characteristics in plane stress conditions and stress-lifetime curves with pre-weathering were also investigated. Finally, a new model predicting the lifetime under the simultaneous application of fatigue loading and weathering-induced degradation was developed by modifying the well-known Miner’s rule.

Key words Weathering; Essential work of fracture; Fatigue lifetime; Polypropylene; Polymer degradation; Lifetime prediction model

1. Introduction Outdoor applications of polymeric materials often cause combined degradation, also known as weathering, due to the joined effect of the sunlight, moisture, heat, biological growth and more [1-4]. Particularly, due to its low cost and adequate performance, polypropylene (PP) with functional fillers like talc is a widely used polyolefin for polymer films, plastic containers, and other molded applications [5-8]. Because of the tertiary carbon in PP monomers, however, free radicals can be easily formed by the abstraction of tertiary hydrogen. Thus, the initiation step of auto-oxidation is promoted, and consequently, the PP and PP-based blends generally reveal a low stability against oxidative deterioration [9, 10]. Such degradation phenomena of the mechanical properties of polymers due to the surrounding chemicals can be distinguished by the presence of a chemical reaction. When the plastics are in contact with aggressive chemicals, stress corrosion cracking (SCC) is caused by chemical deterioration, such as oxidation, and the embrittlement that is due to the chain scission is observed [11, 12]. On the other hand, polymeric materials also can be degraded physically without oxidation. When the contact agents, particularly oils, are diffused into the polymers under mechanical stress, the physical bonding between the adjacent chains could weaken. Thus, the sliding motion between the polymer chains are easier, and the total lifetime would be shortened compared to the inert condition. This occurrence is called environmental stress cracking (ESC) [13, 14]. The weathering of polymeric materials, meanwhile, can be thought of as a combined deterioration process. Due to the increasing usage as load-bearing parts of polymers including PP as exterior structures, an improvement of the product’s reliability through the precise lifespan prediction under weathering is essential. There have been many studies on the effect of weathering on PP based materials. Weathering-induced degradation primarily occurs by photo-oxidation due to the sunlight, which generally accompanies the structural and morphological changes. One of the methods to quantify the degree of weathering-induced degradation is a usage of the carbonyl index (CI),

which is obtained from the FT-IR spectra [1, 3, 15-18]. In the termination step of the autooxidation process of polyolefins, carbonyl function groups (-C=O) are generated as the residues. Because the amount of such carbonyl groups increases with the extent of oxidative degradation, the degradation degree can be evaluated by measuring these residues [11]. These carbonyl groups are detected in the wavelength range of 1600-1800 cm-1 in the FT-IR spectra, and the quantitative term representing the carbonyl peaks is carbonyl index (CI). Leong et al. investigated the variation of the FT-IR spectra and DSC characteristics that resulted from the natural weathering of several PP-based blends [4]. The increase of CI with the natural weathering time was confirmed. They revealed also that the hybrid composites of PP, talc, and calcium carbonate exhibits an enhanced resistance compared to the single-filler PP composites. Rabello et al. investigated the weathering-induced degradation on the PP and talc composites that contained weld-lines [3]. The effect of the aspect ratio of the fillers on the weld-lines and a decrease in adhesion due to weathering were observed. Shyichuk et al. studied the chain scission and chain crosslinking concentration during UV exposure, and revealed that the chain scission is the main mechanism in PP surface layers during photo-degradation [10]. In order to design polymer-based products for outdoor applications, the understanding of the changes of short- and long-term mechanical properties with weathering is also required. Many studies have investigated the variations of tensile properties with weathering-induced degradation [1, 3, 4, 15, 17-20]. Generally, oxidative degradation of polymers causes a decrease in molecular weight and a slight increase in crystallinity, which results in the densification of the surface layer. These kinds of structural changes lead to the slight increase of modulus and strength and decrease in the strain at the break. However, there is a lack of investigations on the changes in fracture properties and fatigue lifetime with regard to weathering-induced degradation. Considering the widespread usage of PP based blends and low stability against oxidation, the comprehensive discussion on the fracture and fatigue properties with weathering is necessary for the reliable design of the PP

based structure [21]. The conventional fracture toughness test such as ASTM E339 reveals the resistance of the fracture in a plane strain condition, where the fracture toughness is independent of the specimen thickness [22]. However, the fracture characteristics in the plane stress state are also imperative because a large amount of exterior plastic materials are usually applied to the sealing or covering of parts in the form of thin sheets. In order to characterize the fracture properties in such plane stress conditions, the essential work of fracture (EWF) test protocol has been usually applied, especially on polymeric materials [23-26]. Because of its straightforward principles and efficient test procedures, the EWF test has been widely accepted to estimate the fracture toughness and resistance to crack growth. Also, the database for fatigue lifetime with regard to the degree of weathering-induced degradation is required for the longterm design of structural components in outdoor applications. In this study, PP and talc blends were exploited to investigate the effect of weathering on the tensile, fracture, and fatigue properties. First, the PP and talc composites were degraded through a custom-made accelerated weathering chamber, and the degradation degree was quantified by CI through the FT-IR spectra. The degradation of the PP blends that undergo field weathering was also measured, and the empirical model of the acceleration factor was constructed by comparing two kinds of weathered specimens. Secondly, the tensile, EWF, and fatigue properties were evaluated with regard to the accelerated weathering time. Finally, a newly suggested fatigue lifetime prediction model, when the fatigue loading and weathering are applied simultaneously, was developed by using the linear damage accumulation concept. Because the most applied scenario in outdoor environments involves mechanical loading and chemical degradation concurrently, the developed model would be beneficial for reasonable lifetime estimation. 2. Materials and methods 2.1 Materials

All tests in this study were performed with the injection molded PP and talc blends, where the weight compositions of PP and talc were 80%, and 20%, respectively. The dumbbell-shaped tensile specimen and rectangular sheets with a dimension of 300 mm × 100 mm were made by an injection molded process, where the thickness of these tensile specimens and sheets was 2.7 mm. The injection molded sheets were cut by a portable band saw to prepare the deep doubleedge notched tensile (DDENT) specimens with a dimension of 100 mm × 30 mm × 2.7 mm. Also, the square plaques (30 mm × 30 mm × 2.7 mm) were machined for the quantification of accelerated and field weathering-induced degradation. All cut sections that were formed by the band saw were polished by sandpaper to a grit size of 800.

2.2 Accelerated and outdoor weathering process Tensile, DDENT, and square specimens were weathered by a custom-made accelerated weathering machine (Fig. 1). In the accelerated weathering process, the UV light was irradiated on the upper surface of the specimens. To achieve the similar light characteristics with sunlight, the commercial Zenon-arc ramp made by USHIO (Model number of UXL-3000HK-O) was employed. The specimens were put on a supporting round plate which rotated at a constant angular speed of roughly 0.2 rad/s for uniform irradiation. The distribution of UV irradiance on the support was measured by a portable irradiance meter. The irradiances at the center and edge were measured at 69 and 45 W/m2, respectively. At the middle radius of the supporting plate, the irradiance was measured at about 57 W/m2 (Fig. 2). The chamber temperature and relative humidity during the accelerated weathering were 85°C and 95%, respectively. In order to simulate the natural weathering process, all accelerating factors were applied in a cyclic manner (Fig. 3). Basically, the 50 hours was set as one cycle time, and 1 hour was applied at both

ends to remove the thermal shock effect. During the accelerated weathering process, all specimens were inverted manually at each cycle to avoid warping to one side. A set of the square plaques was also located on the rooftop of a school building in Seoul, Republic of Korea, for the outdoor weathering process. This outdoor weathering was conducted from July 2016 to April 2017. For the further data, the outdoor-weathered samples for 6 and 9 years were also supplied and analyzed.

2.3 Quantification of the weathering-induced degradation The oxidation process of the polymeric materials can be divided into three parts, i.e., initiation, propagation, and termination stages. Particularly, at the termination step, it is well known that the carbonyl functions (-C=O) are generated as the end products. Thus, the quantification of oxidative weathering can be conducted by measuring the quantity of the amount of such products. In this regard, the degree of degradation from the accelerated and outdoor weathering samples were quantified relatively using the carbonyl index (CI) and assessed by the FT-IR spectrometer with an ATR type (PerkinElmer, Spectrum 100), at the surface of the specimens. In this study, the CI was set as the sum of the absorbance of a,

-

unsaturated ketone (1698 cm-1), acid (1715 cm-1), and ester (1738 cm-1), and normalized by the absorbance at a wavenumber of 2722 cm-1, which is related to the PP-stretching motion. The absorbance at the stretching motion is known as almost invariant with oxidative degradation. Thus, the CI was determined by the following equation,

CI=

A1698cm-1 +A1715cm-1 +A1738cm-1 A2722cm-1

,

(1)

where the An cm 1 denotes the absorbance at the wavenumber of n cm-1. The CI does not indicate the amount of a certain carbonyl functions. However, it can be a useful parameter that

represents global degradation due to the oxidation, so that the quantitative comparison of the degradation degree is possible [11, 16]. Comparing the CIs of the specimens under accelerated and outdoor weathering process, the accelerating factors could be determined empirically.

2.4 Tensile test In order to characterize the tensile properties with accelerated weathering, the dumbbell shape specimens following the type 13B in the KS (Korea Industrial Standards) B 0801 standard, were manufactured by an injection molding [27]. The gage length, width, and thickness of the specimen were 50 mm, 12.5 mm, and 2.7 mm, respectively. The tensile tests were performed with the specimens which underwent the accelerated weathering process for 0, 100, 200, and 400 hours, that is 0, 2, 4, and 8 cycles, respectively. The crosshead speed was 5 mm/min in the room temperature. Five specimens were used for the same accelerated weathering time (tacc) and the average values for the tensile properties were adopted. The elastic modulus (E) was determined by linear fit in a range between 0.0005 and 0.0025 of the engineering strain ( ). The maximum engineering stress and engineering strain at failure were considered the tensile strength (

TS

) and strain at break (

b

), respectively.

2.5 Essential work of the fracture (EWF) test The basic concept of the essential work of fracture (EWF) was originally proposed by Broberg [28] and the specific methodology was established by Cotterell and Reddel [29]. The simple procedures of the EWF test have led to considerable application of this method to characterize the fracture toughness and crack growth resistance in plane stress conditions. First, the total required energy in order for the deep double notched tensile (DDENT) specimen to be fractured into two parts was measured. It should be equal to the area under the loaddisplacement curve. Such total energy (Wf) can be separated into the essential work of fracture (We), which indicates the required energy to generate newly cracked faces, and non-essential

work of fracture (Wp), which signifies the energy dissipation for building the plastic zone just before the actual fracture. Such energy separation can be described by following equation,

Wf

We Wp .

(2)

Because the We is directly related to the crack growth and final fracture, it would be proportional to the initial ligament length (L) of the DDENT specimen as displayed in Fig. 4. In addition, the required energy for the plastic zone that is formed before the actual fracture may be proportional to the volume of the plastic zone, i.e.,

Wf

wp L2t ,

we Lt

(3)

where the we and wp stands for the specific essential work of fracture and specific non-essential work of fracture, respectively. L is the initial ligament size, and t is the thickness of the DDENT specimen.

refers to the shape factor of the plastic zone. The above equation can be

rewritten as:

Wf Lt

wf

we

wp L ,

(4)

where the wf is the specific total work of fracture. As exhibited in this equation, the we and

wp

indicate the intercept and slope of the wf versus L diagram, respectively. Therefore, we and

wp

can be obtained by several EWF tests for various initial ligament sizes, and linear fitting of the wf - L data. Fundamentally, the we implies the fracture toughness of the testing materials under plane stress conditions, which means the crack initiation resistance, and

wp is related to the

dependency of the energy required on the crack length, i.e., the resistance to the crack growth. To verify the EWF properties with regard to accelerated weathering, the machined sheets with a dimension of 100 mm × 30 mm × 2.7 mm were aged through the accelerated weathering process for 0, 100, 200, and 400 hours. After preparing the weathered sheets, deep notches were made in both edges using a razor blade. The ligament lengths (L) of the DDENT specimens were 9, 12,

and 15 mm (Fig. 4). All EWF tests were performed at room temperature and the crosshead speed was 50 mm/min, which is in accordance with the European Structural Integrity Society (ESIS) protocols [30]. 2.6 Fatigue test To investigate the effect of accelerated weathering on the fatigue lifetime of the present PP and talc composites, fatigue tests were conducted in the tension-tension load at room temperature, with the same specimen as the tensile test (KS-13B). The ratio of the minimum and maximum load, known as the R-ratio, was set as 0.1. Compressive stress was excluded for this study to avoid any unexpected buckling phenomenon by the use of thin specimens. The maximum stress levels were in the range of 70~95% of the tensile strength. The frequency of sinusoidal cyclic load was 5 Hz. In the same manner as the tensile and EWF tests, the accelerated weathering samples for 0, 100, 200, and 400 hours were prepared, and the stress-lifetime (S-N) curves regarding the accelerated weathering time (tacc) were obtained empirically.

3. Results and Discussion 3.1 Weathering-induced degradation and accelerating factor Fig. 5a illustrates the FT-IR spectra of the PP and talc composites with regard to the accelerated weathering time (tacc). The magnified view representing the wavenumber range of 1800-1600 cm-1 is also displayed in Fig. 5b. As expected, the carbonyl peaks increased with the tacc. It is worth noting that among the several carbonyl functions, such as

, -unsaturated

ketone, acid, lactone, and more, the ester peak represented by the wavenumber of 1738 cm -1 dramatically soared with accelerated weathering. Furthermore, there was a slight increase of the carbonyl peaks from the initial to the 300 hours of tacc; however, the ester peak rapidly increased between 300 and 400 hours. Fig. 6a displays the FT-IR spectra with the outdoor weathering time (tod); while Fig. 6b illustrates the magnified view for the range of carbonyl peaks. Similar to accelerated weathering, the carbonyl peaks increased with the tod. One can recognize that there was an increase of the absorbance peak at the wavenumber of 1640 cm -1, which indicates the formation of a vinyl group. Whereas, the peaks around 1640 cm -1 were almost consistent in the case of accelerated weathering (Fig. 5b). The formation of the vinyl species in outdoor weathering of PP-based materials has been reported in several studies [2, 4]. The increase of the carbonyl index (CI) with tacc and tod is revealed in Fig. 7a and 7b, respectively. In the case of the accelerated weathering process, the CI increased weakly up to 200 hours, and it rapidly increased after such weathering time of accelerated degradation. Thus, it can be speculated that there is a required duration to initiate sufficient oxidative degradation. Such a period can be thought as the oxidation induction time (OIT) because of the existence of anti-oxidants (AO) and other stabilizers in the samples [12]. In the case of the polymeric materials with AO additives, the oxidation of the base materials would proceed after the depletion of the AO by the oxidation itself. Because the materials to be oxidized during the OIT and after the OIT are different, the variation of the CI with the tacc should be modeled by

different functions with different regions. In this study, the parabolic and dose-response functions were employed to conduct the fitting before the OIT, and after the OIT, respectively. The fitting equations are as follows:

CI tacc

0.438 0.00931 tacc

CI tacc

1.037

2.31 10

tacc

2

for 0 tacc

200 h (5)

7.286 1 10

5

0.007 387.1 tacc

for tacc

200 h.

Moreover, by matching the CI at the accelerated (Eq. 5) and outdoor weathering processes, the following accelerated weathering factors can be also constructed (Fig. 7c).

tacc tacc

0.00385 tod

8.0638 10

309.17 0.00233 tod

6

for tod

tod

2

for 0

tod

6570 h

(6)

6570 h,

where the unit of tacc and tod is in hours. It should be noted that the Equations (5) and (6) are purely empirical equations, and these are valid only for certain weathering conditions described in the present study. In addition, the sample surfaces were observed with regard to the tacc through an optical microscope (Fig. 8). Until 200 hours of the tacc, there is no significant difference of the surface morphology with the tacc; however, at 400 hours, heterogeneous whitening regions due to chemical degradation were observed. It arose from the densification caused by the chain scission. Such morphological changes were in good agreement with the rapid increase in CI after the 200 hours of tacc, which indicates that the serious deterioration of the PP chains occurs after the OIT.

3.2 Tensile properties The stress-strain curves for the PP and talc composites under the tensile test, with different accelerated weathering times (tacc), are provided in Fig. 9a. Generally, the oxidative degradation on the thermoplastics results in the chain scission and embrittlement, which leads to the subtle increase in the elastic modulus (E) and tensile strength (

TS

), and reduction in the ductility. As

expected, the elastic modulus and tensile strength slightly increased with the tacc. It is well known that the chain scission, due to the oxidative degradation, leads to the ease of the packing of molecular chains together, which results in an increase in the crystallinity [31]. The increase in the crystallinity perhaps caused an increment in the modulus and tensile strength (Fig. 9b). The loss of ductility due to the degradation-induced chain scission was manifested in the dramatic decrease in the strain at the break (

b

) (Fig. 9c). The normalized variations of the

mechanical properties, by those initial values, are given in Fig. 9d. It can be clearly seen that the strain at the break, one of the parameters representing the ductility, mostly changed due to the oxidative degradation rather than the tensile strength and elastic modulus. Such trends are in accordance with the previous research regarding the effect of oxidative degradation on thermoplastics [4]. 3.3 Essential work of fracture (EWF) properties The load (P) – displacement (

) curves of the DDENT specimen for different initial

ligament sizes (L) and different accelerated weathering times (tacc) are exhibited in Fig. 10. As expected, the larger the L that the DDENT specimen has, the more

and total energy it

requires to be fractured. Such required energy (Wf) can be calculated by integrating the area under the P-

curves. Fig. 11a reveals the dependence of the specific work of fracture (wf) on

the L, where the wf is defined as Wf divided by the initial ligament area, Lt. As described in Equation (4), the specific essential work of fracture (we) and the specific non-essential work of fracture (βwp) are determined by the intercept and slope of the wf - L curves, respectively. The

variation of the we and βwp with regard to the tacc is illustrated in Fig. 11b. First, the we, which indicates the fracture toughness in a plane stress condition, decreased with the tacc, in the early region. However, the we decreased with the tacc after 200 hours of degradation time. In other words, the fracture toughness, which represents the resistance to the crack initiation, increased with regard to the oxidative degradation degree. It seems that the increase in tensile strength (

TS

) due to the embrittlement may be a main cause of such phenomena. Second, the βwp, which

represents the required energy density to form the plastic zone, reduced consistently with the degradation. This result implies that the plasticity of the materials decreased due to the reduction in the ductility, which manifested in a drop of the strain at the break (

b

) with tacc.

From a different viewpoint, the slope of the wf – L diagram can be considered the dependence of the surface fracture energy, which means the resistance to generating new crack faces, on the crack length. Therefore, it can be explained that the resistance to the crack propagation also reduced with the oxidative degradation. 3.4 Fatigue tests The stress-lifetime (S-N) diagram with regard to the accelerated weathering time (tacc) is given in Fig. 12a. The vertical and horizontal axes represent the stress amplitude (

a

), half of

the stress range, and logarithmic cycles to failure (Nf), respectively. During the early stages of weathering, until the 100th hour, the lifetime somewhat increased at the high stress amplitude levels, and decreased at the low stress amplitude levels. Those tendencies manifested into the increase in the slope of the S-N diagram. At 200 hours of tacc, the S-N curve shifted markedly to the right, which indicates an increase in fatigue lifetime. Similar to the EWF results, the increase in the elastic modulus (E) and tensile strength (

TS

) with the tacc induces a longer fatigue

lifetime in the early weathering period. However, after further degradation, the S-N curve shifted to the left again, with a decreased slope, which indicates a decrease in the fatigue lifetime with the tacc. It indicates that the weathering-induced degradation had an adverse effect on the long-

term mechanical performance, which was represented by the fatigue lifetime; while short weathering affects the samples rather favorably. The experimental S-N curves were fitted with the below empirical equation;

a

A log10 N f

(7)

B,

where the A and B indicate the slope and intercept in the S-N diagram, respectively. The fitted curves are shown as dotted lines in Fig. 12a. The variations of A and B with the tacc are represented in Fig. 12b, and each parameter was fitted by the parabolic equation;

A

a0

a1tacc

a2 tacc 2

B

b0

b1tacc

b2 tacc 2 ,

(8)

for the tacc in hours. The coefficients in Equation (8) is provided in Table 1. Due to the competition between strengthening/stiffening and embrittlement of the material by weathering, it is expected to have two coefficients (A and B) in non-linear (rather parabolic) relationship with weathering time (tacc). The crosshead displacement at peak stress during the fatigue,

a

peak

with several tacc and

levels are depicted in Fig. 13. There were three kinds of development of the

regard to the tacc and peak

a

. For the undegraded samples at the high

a

peak

with

, a smooth increase of the

was observed with the fatigue procedure, and the increase in the

peak

the fatigue failure and after the constant rate region (Fig. 13a). At the moderate

rate just before

a

and shortly

degraded samples, the fatigue failure occurred after the constant rate region without the rapid increase in the rate of the

peak

(Fig. 13b). As displayed in Fig. 13c, for the further degradation

and low stress level, there were several stepwise jumps in the

peak

, and the peak displacement

at failure was lower than Fig. 13a and 13b. These distinctive fatigue failure modes were referred to as ductile, brittle, and brittle-step. In the case of the ductile mode, it did not imply the well-

known ductile failure, but did express the relatively ductile mode because the higher value of the peak

was required for the ultimate failure than other modes. Fig. 14 displays the corresponding

fracture surface for the cases in Fig. 13. There were no remarkable differences in the fracture surfaces, so the fatigue fracture modes were distinguished by the aspect of the

peak

modes. The fracture mechanism map (FMM) in the fatigue tests regarding the tacc and

growth

a

is

depicted in Fig. 15. At the higher mechanical stress and lower degree of degradation, the fatigue failure occurred in the ductile mode; whereas, specimens failed in the brittle-step mode at the lower applied stress and high degree of degradation. At the intermediate region, the specimens that were under the cyclic load fractured in the brittle or brittle-step mode. 3.5 Development of the fatigue lifetime prediction model under simultaneous weathering-induced degradation The empirical fatigue lifetime shown in Fig. 12 has a limitation to the practical lifetime expectation because these results are lifetime under fatigue loading of pre-weathered samples. In other words, purely mechanical stress was applied following a certain level of weatheringinduced degradation. However, in the real service conditions, the outdoor polymeric materials usually underwent mechanical stress and weathering simultaneously; therefore, the development of the lifetime prediction model under such concurrent effects is required to estimate the lifetime appropriately. The estimation of fatigue lifetime under various stress amplitudes can be performed by the well-known Miner’s rule, using the linear damage accumulation concept [32]. In this part, the Miner’s rule will be modified for the present study. First, there are two applied terms to deteriorate the samples in this study, the stress amplitude (

a

) for the mechanical component

and accelerated weathering time (tacc) for the weathering one. Other factors affecting the lifetime such as the R-ratio, fatigue frequency, and accelerated weathering temperature were kept consistent. To predict the lifetime under the combined condition, a small working-time

increment,

t , was set. As displayed in Table 2, at the nth step, the working-time of the

considered component becomes n t . Then, the average accelerated weathering time at nth step becomes

2n 1 t / 2 . During a time interval, the applied cycles,

N , can be directly

calculated by f t , where f is the loading frequency. The corresponding lifetime at the condition of each step (Nf, n) can be also determined by the empirical relationship, Equations (7) and (8). By using the linear damage accumulation concept, the damage fraction at each step can be calculated as

N / N f ,n . As the operating time goes by, the damage fractions would be

superposed, and final failure occurs providing that the accumulated damage fraction n

N / N f ,i becomes a unit value. In the case of the fatigue loading frequency of 5 Hz, which is i 1

the same frequency as the experiments, the predicted S-N curve under the simultaneous conditions is depicted as a solid curve in Fig. 16a. Understandably, at the high

a

region, the

predicted lifetime is similar to those of 0 and 100 hours, because the weathering period is too short for the samples to undergo enough weathering-induced degradation. At the intermediate a

level, it comes closer to the lifetime for the tacc of 200 hours, and a further decrease in

a

leads to ensuring a sufficient weathering period, which results in the approach to the S-N curve for 400 hours of tacc. For further discussion, the effect of loading frequency on the suggested model is also represented in Fig. 16b. Because the weathering process is related to the operating time, not the number of cycles, the influence of weathering should be reduced with an increase in loading frequency, and vice versa. The effect of fatigue frequency should be converted into the time scale, but weathering effect is a direct function of time. At a high frequency of 20 Hz, the effect of weathering on the total lifetime is relatively small compared to the mechanical stress. Thus, the predicted S-N curve in this case is similar to that of undegraded samples (0 h). However, at the low loading frequency of 0.2 Hz, the time to be degraded by weathering can be retained

sufficiently before the failure, and the effect of weathering within the total lifetime would increase; while the dependence of the lifetime on the be clearer at the low

a

a

would decrease. Such a trend should

region, and the dramatic reduction in lifetime with decreasing loading

frequency would be easily identified. In consequence, it can be demonstrated that the developed model to estimate the lifetime under the simultaneous effect of mechanical loading and weathering exhibits the appropriate tendencies, and this methodology can be used practically in the design for the outdoor application of polymeric materials. The effect of frequency should be explored in detail with some additional experimental data in the future.

4. Conclusions In this study, the effect of weathering-induced degradation of polypropylene (PP) and talc composites on the short-term and long-term mechanical performances were examined. The degradation degrees of accelerated and outdoor weathered samples were assessed quantitatively using FT-IR spectroscopy, and the accelerated weathering factor was constructed by comparing the carbonyl index (CI) of the weathered samples. The elastic modulus (E) and tensile strength (

TS

) slightly increased with regard to the accelerated weathering time (tacc);

while the strain at the break (

b

) dropped dramatically, which indicates embrittlement due to

weathering. Through the essential work of fracture (EWF) methodology, the fracture properties in plane stress conditions were also characterized. The specific essential work of the fracture, we, which represents the plane stress fracture toughness, increased with the tacc following a short deceasing period of early weathering. Such a result arose from the combined effect of the increase in modulus and strength, and decrease in toughness. The specific non-essential work of fracture, βwp, which indicates the energy dissipation density in the plastic zone, consistently decreased with tacc, due to the embrittlement of the degraded specimens. The stress-lifetime (S-

N) curves with tacc were constructed. In the early degradation period, the S-N curve shifted right, which indicates a subtle increase in lifetime due to the strengthening and stiffening of the samples with the oxidative degradation. However, the longer the weathering time was applied, the shorter of a lifetime the samples had, owing to the significant degradation. Finally, by modifying the well-known Miner’s rule, a fatigue lifetime estimation model for the simultaneous application of both cyclic loading and weathering was developed. The suggested model exhibited the proper behavior and the effect of fatigue loading frequency was also simulated reasonably. The developed model provides a more realistic lifetime-assessment methodology for plastic components working in actual outdoor conditions.

Acknowledgement This work was supported by the Technology Innovation Program Project (No. 10076562) of Korea Evaluation Institute of Industrial Technology (KEIT) funded By the Ministry of Trade, industry & Energy (MI, Korea).

References [1] A. Tidjani, Photooxidation of polypropylene under natural and accelerated weathering conditions, Journal of Applied Polymer Science, 64 (1997) 2497-2503. [2] Y. Lv, Y. Huang, J. Yang, M. Kong, H. Yang, J. Zhao, G. Li, Outdoor and accelerated laboratory weathering of polypropylene: A comparison and correlation study, Polymer Degradation and Stability, 112 (2015) 145-159. [3] M. Rabello, R. Tocchetto, L. Barros, J. d'Almeida, J. White, Weathering of polypropylene composites containing weldlines, Plastics, rubber and composites, 30 (2001) 132-140. [4] Y. Leong, M.A. Bakar, Z.M. Ishak, A. Ariffin, Characterization of talc/calcium carbonate filled polypropylene hybrid composites weathered in a natural environment, Polymer Degradation and Stability, 83 (2004) 411-422. [5] O. Ammar, Y. Bouaziz, N. Haddar, N. Mnif. Talc as reinforcing filler in polypropylene compounds: effect on morphology and mechanical properties. Polymer Science, 3 (2017) 1-8.

[6] R. Nasrin, M.A. Gafur, A.H. Bhuiyan. Characterization of isotactic polypropylene/talc composites prepared by extrusion cum compression molding technique. Materials Sciences and Applications, 6 (2015) 925-934. [7] A.M. Hartl, M. Jerabek, R.W. Lang. Anisotropy and compression/tension asymmetry of PP containing soft and hard particles and short glass fibers. Express Polymer Letters, 9 (2015) 658670. [8] A.O. Bouakkaz, A. Albedah, B.B. Bouiadjra, S.M.A. Khan, F. Benyahia, M. Elmeguenni. Effect of temperature on the mechanical properties of polypropylene-talc composites. Journal of Thermoplastic Composite Materials. 31 (2018). [9] A. Shyichuk, D. Stavychna, J. White, Effect of tensile stress on chain scission and crosslinking during photo-oxidation of polypropylene, Polymer Degradation and Stability, 72 (2001) 279-285. [10] A. Shyichuk, J. White, I. Craig, I. Syrotynska, Comparison of UV-degradation depth-profiles in polyethylene, polypropylene and an ethylene–propylene copolymer, Polymer Degradation and Stability, 88 (2005) 415-419. [11] B.-H. Choi, A. Chudnovsky, R. Paradkar, W. Michie, Z. Zhou, P.-M. Cham, Experimental and theoretical investigation of stress corrosion crack (SCC) growth of polyethylene pipes, Polymer Degradation and Stability, 94 (2009) 859-867. [12] B.-H. Choi, A. Chudnovsky, K. Sehanobish, Stress corrosion cracking in plastic pipes: Observation and modeling, International Journal of Fracture, 145 (2007) 81-88. [13] B.H. Choi, J. Weinhold, D. Reuschle, M. Kapur, Modeling of the fracture mechanism of HDPE subjected to environmental stress crack resistance test, Polymer Engineering & Science, 49 (2009) 2085-2091. [14] J.-W. Wee, Y. Zhao, B.-H. Choi, Observation and modeling of environmental stress cracking behaviors of high crystalline polypropylene due to scent oils, Polymer Testing, 48 (2015) 206214. [15] M. Rabello, J. White, Photodegradation of talc‐filled polypropylene, Polymer composites, 17 (1996) 691-704. [16] H. Nakatani, H. Shibata, K. Miyazaki, T. Yonezawa, H. Takeda, Y. Azuma, S. Watanabe, Studies on heterogeneous degradation of polypropylene/talc composite: Effect of iron impurity on the degradation behavior, Journal of Applied Polymer Science, 115 (2010) 167-173. [17] J.S. Fabiyi, A.G. Mcdonald. Degradation of polypropylene in naturally and artificially weathered plastic matrix composites. Maderas. Ciencia y tecnologia, 16 (2014) 275-290. [18] Y. Lv, Y. Huang, J. Yang, M. Kong, H. Yang, J. Zhao, G. Li. Outdoor and accelerated laboratory weathering of polypropylene: a comparison and correlation study. Polymer Degradation and Stability, 112 (2015) 145-159. [19] B. Fayolle, L. Audouin, J. Verdu, Oxidation induced embrittlement in polypropylene—a tensile testing study, Polymer Degradation and Stability, 70 (2000) 333-340. [20] I. Yakimets, D. Lai, M. Guigon, Effect of photo-oxidation cracks on behaviour of thick polypropylene samples, Polymer Degradation and Stability, 86 (2004) 59-67. [21] M.M. Shokrieh, M. Esmkhani, F. Taheri-Behrooz. A novel model to predict the fatigue life of

thermoplastic nanocomposites. Journal of Thermoplastic Composite Materials. 28 (2015) 14961506. [22] ASTM E399-12e3, Standard Test Method for Linear-Elastic Plane-Strain Fracture Toughness KIc of Metallic Materials, ASTM International, West Conshohocken, PA, (2012). [23] T. Bárány, T. Czigány, J. Karger-Kocsis, Application of the essential work of fracture (EWF) concept for polymers, related blends and composites: A review, Progress in Polymer Science, 35 (2010) 1257-1287. [24] B.-H. Choi, M. Demirors, R.M. Patel, A. Willem deGroot, K.W. Anderson, V. Juarez, Evaluation of the tear properties of polyethylene blown films using the essential work of fracture concept, Polymer, 51 (2010) 2732-2739. [25] J.-M. Lee, B.-H. Choi, J.-S. Moon, E.-S. Lee, Determination of the tear properties of thermoplastic polyester elastomers (TPEEs) using essential work of fracture (EWF) test method, Polymer Testing, 28 (2009) 854-865. [26] A. Martinez, J. Gamez-Perez, M. Sanchez-Soto, J.I. Velasco, O. Santana, M.L. Maspoch, The essential work of fracture (EWF) method–analyzing the post-yielding fracture mechanics of polymers, Engineering Failure Analysis, 16 (2009) 2604-2617. [27] KS B 0801, Test pieces for tensile test for metallic materials, Korea Agency for Technology and Standards, Seoul, Republic of Korea, (2007). [28] K. Broberg, On stable crack growth, Journal of the Mechanics and Physics of Solids, 23 (1975) 215-237. [29] B. Cotterell, J. Reddel, The essential work of plane stress ductile fracture, International Journal of Fracture, 13 (1977) 267-277. [30] Y. Mai, B. Cotterell, Effect of specimen geometry on the essential work of plane stress ductile fracture, Engineering Fracture Mechanics, 21 (1985) 123-128. [31] M. Rabello, J. White, Crystallization and melting behaviour of photodegraded polypropylene—I. Chemi-crystallization, Polymer, 38 (1997) 6379-6387. [32] M.A. Miner, Cumulative damage in fatigue, J. Appl. Mech., 12 (1945) A159-A164.

Table 1 Coefficients for parabolic equations, Eq (8). Coefficients

Values

a0

0.68733

a1

0.00208

a2

-3.681E-6

b0

12.662

b1

0.01335

b2

-2.5963E-5

Table 2 Schematics of fatigue life prediction model under the mechanical cyclic loading and accelerated weathering simultaneously.

Step i

Working time

Time interval

Average tacc

Applied cycles

Corresponding lifetime

A da

1

Δt

0 ~Δt

Δt/2

ΔN

Nf,1=Nf(σa , Δt/2)

ΔN

2

2Δt

Δt ~2Δt

3Δt/2

ΔN

Nf,2=Nf(σa , 3Δt/2)

ΔN +

3

3Δt

2Δt ~3Δt

5Δt/2

ΔN

Nf,3=Nf(σa , 5Δt/2)

ΔN + +

:

:

:

:

:

:

:

n

nΔt

(n-1)Δt ~nΔt

(2n-1)Δt/2

ΔN

Nf,n=Nf(σa , (2n-1)Δt/2)