The determining role of calcium carbonate on surface deformation during scratching of calcium carbonate-reinforced polyethylene composites

Materials Science and Engineering A 404 (2005) 208–220

The determining role of calcium carbonate on surface deformation during scratching of calcium carbonate-reinforced polyethylene composites M. Tanniru a,b , R.D.K. Misra a,b,∗ , K. Berbrand c , D. Murphy c a

Center for Structural and Functional Materials, University of Louisiana at Lafayette, P.O. Box 44130, Lafayette, LA 70504-3130, USA b Department of Chemical Engineering, University of Louisiana at Lafayette, P.O. Box 44130, Lafayette, LA 70504-3130, USA c Westlake Polymers Corporation, P.O. Box 3608, Sulphur, LA 70664, USA Received in revised form 18 May 2005; accepted 18 May 2005

Abstract The scratch-induced surface damage of neat and calcium carbonate-reinforced high density polyethylene is described in terms of characteristics of scratch morphology and scratch deformation parameters. Under identical test conditions, calcium carbonate-reinforced high density polyethylene composites exhibit significantly reduced susceptibility to scratch deformation and stress whitening compared to neat high density polyethylene. The resistance to scratch deformation is discussed in terms of tensile modulus, elastic recovery, and scratch hardness. © 2005 Elsevier B.V. All rights reserved. Keywords: Polyethylene composites; Microstructure; Scratch deformation

1. Introduction Thermoplastic materials are being increasingly considered in automotive, aerospace, electronic, and optical applications where the resistance of the material to surface damage is of particular significance. Thermoplastics such as polypropylene and polyethylene have excellent chemical resistance and have emerged as ideal materials for automobile and consumer products. However, the susceptibility of thermoplastics to surface damage such as scratch is a serious concern [1–5]. Scratch introduces visual damage in materials and may also act as a macroscopic stress raiser reducing the mechanical strength [5–10]. Also, plastically deformed polymeric materials exhibit a whiter appearance called ‘stress whitening’ that is detrimental to optical clarity and aesthetic perception. It is also detrimental to tensile and fatigue loading because it will imply that the energy absorbing mechanisms under stress has been exhausted; therefore, further stress application will lead to early failure. Thus, the increase in the susceptibility of thermoplastic materials to scratch damage is important for maximizing their applications. ∗

Corresponding author. Tel.: +1 337 482 6430; fax: +1 337 482 1220. E-mail address: [email protected] (R.D.K. Misra).

0921-5093/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2005.05.031

Stress whitening is the scattering of visible light after reflection and is a measure of the visibility of scratch. Visible light has a wavelength of 0.6 ␮m and hence voids in the size range of 0.5–1.0 ␮m or deformed features that reflect white light efficiently will enhance stress whitening. Polymeric materials experience different extent of deformation depending on the applied load and a number of test methods have been developed to simulate the actual conditions. Under practical conditions, objects of different geometries may be responsible for the scratch. Thus, the understanding of scratch pattern produced by indenters of different geometries is important [11,12]. Briscoe et al. examined the influence of applied normal load, sliding velocity, angle of indenter, and lubrication [1,13]. Scratch deformation processes of diverse nature including fully elastic, elastic–plastic, ironing, wedging, crazing–tearing, grooving, edge-cracking, and chipping were observed depending on the applied normal load, cone angle, and scratch velocity. The formation of voids and microcrazing are generally believed to be responsible for the origin of stress whitening during scratch damage in polymeric materials. Studies on the mechanism of crazing have shown that the formation of deformation bands initiate crazing and their growth is perpetuated by the extension of existing voids into the

M. Tanniru et al. / Materials Science and Engineering A 404 (2005) 208–220

bulk polymer, thereby linking up the stretched fibrils in their path [10,14]. Other micromechanisms such as shear deformation bands, kinking, microcrazing, and ductile ploughing have also been observed. Misra and co-workers studied the micromechanisms of stress whitening in relation to tensile deformation behavior [15,16] and strain rate sensitivity effect [17]. On the basis of these studies, strain–strain rate deformation maps elucidating the deformation processes responsible for stress whitening during scratch damage was developed. The relationship between deformation mechanisms and scratch visibility was examined using the phenomena of light scattering and reflection from scratches [3,18–20]. This approach involved the quantification of polarized light scattered from the scratch by considering the difference in intensities between bright and dark regions to compare the scratch visibility of different materials. The quantification of stress whitening sensitivity method was improved by Wang et al. [10,21,22] with the help of optically scanned images. It was noted that the extent and severity of plastic deformation determined the efficiency of light scattering. Hence, severely deformed fibrils involving tearing/breaking were more visible than smooth and well-defined grooves [10,21,22]. The inherent physical and mechanical characteristics of polymeric materials that influence resistance to scratch deformation are percentage crystallinity, molecular weight, modulus, and yield strength. Previous studies [10,23] have shown that a decrease in crystallinity of about 5% obtained by cooling polypropylene at different cooling rates enhanced stress whitening. The reduction in the susceptibility of high crystallinity polymers to stress whitening was related to their high modulus and yield stress. The desirable mechanical properties required to minimize susceptibility to surface damage and stress whitening can be accomplished by increasing crystallinity, use of short chain polymers, compounding with additives, and more importantly by reinforcement of polymer with micro- and nano-size minerals [24–30]. In general, reinforcement with minerals increases the modulus and yield stress, and influences the mobility of molecular chains. Previous studies on polymer composites containing micrometer size mineral particles have proven that a higher tensile modulus is obtained in comparison to unreinforced polymer [24–30]. Also, reinforcement particles such as clay may provide additional nucleation sites, thereby increasing the number of spherulites and reducing their size, with consequent increase or decrease in crystallinity [30]. A wide range of inorganic mineral reinforcements has been investigated in recent years from the viewpoint of improving scratch resistance. Primary among those are wollastonite [16,17,22,24,31–34], talc [31–34], and clay [29,35]. While significant improvement in scratch resistance was observed with these minerals, they, however, experienced a significant loss in ductility and toughness in relation to their neat counterparts. In contrast to wollastonite, talc, and clay, the reinforcement of high density polyethylene with calcium carbonate enhanced the impact strength, a behavior attributed to particle-induced cavitation [36]. In the study

209

described here, a comprehensive effort is made to establish a link between reinforcement-induced structural changes and mechanical properties to surface damage and associated stress whitening. To accomplish this objective, a microstructural approach has been adopted to make a comparative evaluation of neat high density polyethylene and calcium carbonate-reinforced polyethylene processed under similar conditions, ensuring that the reinforcement-induced structural changes and stress whitening associated with surface damage is a true reflection of the reinforcement.

2. Experimental procedure 2.1. Materials Standard ASTM tensile bars (ASTM D-638) and Izod impact bars (ASTM D-256) of neat polyethylene and CaCO3 polyethylene composites (5, 10, and 20%) were injection molded and slowly cooled under quiescent conditions. The longitudinal axis of the tensile and impact specimens corresponded to the molding direction. The median size of CaCO3 was ∼1.2 ␮m. Izod impact samples were notched according to ASTM D-256 with a notch-cutting device. A v-notch with a root radius of 0.1 mm was introduced into the specimen using a v-notch cutter to a depth of 2.5 mm. A notch verification device verified the depth. The nominal melt flow index of neat polyethylene was 20 g/10 min at 230/2.16 (i.e., at 230 ◦ C and 2.16 kg piston force). The number and weight average molecular weights were 14,600 and 55,000, respectively and the polydispersity index was 3.77. 2.2. Structural characterization 2.2.1. Crystallinity and lamellar thickness The study of degree of crystallinity assumes particular significance because higher crystallinity, in general, increases modulus and yield stress, and reduces toughness. The degree of crystallinity and lamellar thickness was measured by Xray diffraction. Samples with a thickness of ∼2 mm were used for the X-ray diffraction measurements with the plane of view being parallel to the injection molding direction. The 1D WAXS detector was first calibrated using the known peak positions of polyethylene for the given wavelength (Cu K␣ radiation = 0.1542 nm) using Bragg’s law: nλ = 2d sin θ

(1)

where n is the order of reflection, λ is the wavelength of incident radiation, d the interplanar spacing, and θ is the scattering angle. The degree of crystallinity, χ, of semi-crystalline sample was determined by: χ=

C−A C

(2)

where C is the area of the crystalline profile, and A is the area of the amorphous profile. The crystalline long period, D, is

M. Tanniru et al. / Materials Science and Engineering A 404 (2005) 208–220

210

related to the lamellae thickness, Lc , by: Lc = χD

(3)

Besides crystallinity measurements by X-ray diffraction, the crystallization behavior was studied by differential scanning calorimetry (DSC), to study the effect of calcium carbonate. The samples were cut across the section and heated from room temperature (∼20 ◦ C) to 180 ◦ C at a rate of 10 ◦ C/min, held at 180 ◦ C for about a minute, cooled to −20 ◦ C at the rate of 10 ◦ C/min, and a second scan was carried out in a manner similar to the first. Using the DSC data, the lamellar thickness (l) is given by the Thomson–Gibbs equation [37]: l=

2γTm◦ Hρ

(4)

where Tm◦ is the equilibrium melting temperature, γ the surface free energy, H the heat of fusion for 100% crystalline polyethylene, and ρ the density. Wide-angle X-ray diffraction data was also used to determine if there was any preferential alignment of reinforcement particles during injection molding.

impact tester (model 899) with an impact velocity of 1 m/s. At least three samples were tested for each condition. 2.4. Scratch tests Surface damage was introduced by producing a welldefined scratch on the surface. The scratch equipment consisted of a balance beam scrape adhesion tool that utilized different types of stylus, namely, Hoffmann, needle, and loop. In this study, a Hoffmann-type stylus having a hardened stainless steel cylinder and a contact arc of diameter 6.945 mm was used. The samples were fixed on a leveling platform attached to a displacement stage and normal load was applied by placing dead-weights on the indenter holder. Mechanically-induced surface damage in the form of a scratch was introduced on the surface of the samples using loads in the range of 2–6 kg. The scratch test conditions were identical for all the four polymeric materials to enable a direct comparative evaluation. The scratch direction was the longitudinal axis of the injection-molded specimen. 2.5. Surface deformation

2.2.2. Spherulitic structure and nucleating efficiency of calcium carbonate The spherulite size was studied by etching the samples with 5% potassium permanganate–sulfuric acid solution at room temperature followed by examination in SEM. In order to determine the nucleating efficiency of CaCO3 , the nucleating efficiency scale proposed by Fillon et al. [38,39] was used. This scale is based on conventional DSC cooling runs and utilizes crystallization range determined in self-nucleation experiments to correlate to spherulite size. The proposed calorimetric efficiency scale approach for nucleating additives is more satisfactory than just comparing peak crystallization temperatures upon dynamic cooling from the melt because it is used as a reference not for the neat polymer but for the polymer that is self-nucleated. The nucleating efficiency is given by [38,39]: N = 100 ×

TCNA − TC TCmax − TC

(5)

where TCNA is the peak crystallization temperature of the polymer with the nucleating agent, TC is the crystallization temperature, and TCmax the optimum self-nucleation temperature. 2.3. Tensile and impact properties Tensile properties (modulus, yield strength) were determined by testing the tensile bars (ASTM D 638) of neat and CaCO3 -polyethylene composites in uniaxial tension at 20 ◦ C using a computerized MTS 210 tensile testing machine at selected displacement rate of 50 mm/min (strain rate of ∼0.017/s). The Izod impact toughness (ASTM D 256) was determined using an instrumented falling weight Tinius Olsen

The microstructural evolution associated with the scratch process was studied using field emission scanning electron microscope (JEOL 6300F). Sections were cut from the scratched area and the characteristics of deformation process and micromechanisms involved were examined. 2.6. Quantification of surface damage The effect of calcium carbonate on resistance to surface deformation was evaluated by atomic force microscopy (AFM, Nanoscope (R) IIIa, Digital Instruments, Santa Barbara, CA) in terms of average scratch roughness. The AFM provides excellent Z-resolution to quantify the surface topography. Atomic force microscopy was carried out using tapping mode to prevent possible surface damage due to continuous contact. All the scans were made in air and the tip used for the study of samples was ‘tapping mode etched silicon probe’ (TESP). The length of the tip was 125 ␮m. The scan parameters and AFM scan leveling procedures were identical for all the investigated polymeric materials to enable direct comparison of images. A number of measurements were made at different positions along the scratch for each sample to ensure that data was a true representative of the characteristics of the scratched surface. To quantify the stress whitening or visual damage associated with the scratch deformation, stress whitened samples of both neat and reinforced polymeric materials were optically examined in the dark field image mode using a Sony CCD camera and the images were saved as tagged image file format (TIFF). Subsequently, the recorded images were processed in the gray mode using Image J software (based on NIH software). The image analysis functions available in the

M. Tanniru et al. / Materials Science and Engineering A 404 (2005) 208–220

software include gray value against x–y coordinates of all the pixels.

3. Results and discussion 3.1. Physical and mechanical characteristics 3.1.1. Crystallinity and macromolecular structure The percentage crystallinity obtained from X-ray diffraction is listed in Table 1. The crystallization data (percentage crystallinity, melting and crystallization temperatures) obtained from DSC experiments for neat and CaCO3 polyethylene are also presented in Table 1. The percentage crystallinity was estimated using a value of heat of fusion of 292 J/g for 100% crystalline polyethylene. A similar increase in percentage crystallinity with CaCO3 is observed from DSC. The increase in bulk crystallinity on reinforcement with CaCO3 is significant and is expected to influence the mechanical behavior (modulus, yield strength) and the scratch deformation behavior. Also, listed in Table 1 are lamellar thickness (l), long spacing (D), spherulite size, and nucleating efficiency of CaCO3 . Representative SEM micrographs of the spherulite morphology of neat polyethylene and 20% CaCO3 polyethylene composites are presented in Fig. 1 (spherulite morphology of 5 and 10% CaCO3 polyethylene are not presented). The spherulites grow from a central nucleating entity and radiate outwards. However, there were differences in the morphology of the spherulite between neat and polyethylene composites. In neat polyethylene, the spherulites were characterized by continuous branching and fanning of lamellae that evolved into sheaves and finally into spherulites, while the growth of spherulite in CaCO3 -reinforced polyethylene composite developed as straight rod-like structure. Also, there was a consistent decrease in spherulite size with increase in spherulite size (Table 1) from ∼400 ␮m in neat polyethylene to 37 ␮m in 20% CaCO3-polyethylene composite. The large decrease in spherulite size implies the strong nucleating effect of CaCO3 reinforcement. Using equation [5], the nucleating efficiency varied from 44% for 5% CaCO3 to 32% for 20% CaCO3

211

polyethylene composite (Table 1). The small increase in efficiency scale with increase in %CaCO3 is consistent with the decrease in spherulite size (Table 1). These observation imply that we have not reached the maximum limit of nucleating efficiency at 20% CaCO3 . The nucleating efficiency scale observed by Fillon et al. [38] for typical nucleating agent (carboxyl acid) was in the range of 60–70%, and is similar to our observations. Additionally, the percentage crystallinity and lamellar thickness followed a similar trend, i.e., a higher percentage crystallinity material was characterized by greater lamellar thickness (Table 1), as expected. Considering the alignment of particles during molding has a strong influence on the mechanical behavior and inevitably contribute to property difference, the X-ray diffraction data was further analyzed besides determination of crystallinity. The X-ray diffraction peak intensity data (Io ) for major (1 0 4) and (2 0 2) planes for hexagonal CaCO3 , and (1 1 0) and (2 0 0) planes in the orthorhombic unit cell for polyethylene are summarized in Table 2. This data is normalized with respect to the most intense (1 1 0) polyethylene peak to eliminate variations between the effects of peak intensity variations in samples. The data in Table 2 shows the following: (a) Normalized intensity of (2 0 0) plane of polyethylene with respect to (1 1 0) is similar in both neat and 20% CaCO3 -reinforced polyethylene (neat = 0.33, composite = 0.39) and is similar to the standard I110 /I200 value of ∼0.5 [36]. (b) The experimental I(104) /I(202) peak intensity ratio (∼6) for the CaCO3 in the reinforced composite is similar to the standard I(104) /I(202) value of 5 [36] for CaCO3 .

Table 2 X-ray diffraction data for neat high density polyethylene (PE) and calcium carbonate-reinforced polyethylene compsoites Material

PE 20% CaCO3 –PE

Normalized peak intensity data PE (I110 )

PE (I200 )

CaCO3 (I104 )

CaCO3 (I202 )

1.0 1.0

0.33 0.39

– 1.06

– 0.17

Table 1 Physical and mechanical properties of high density polyethylene calcium carbonate-reinforced polyethylene composites Material

Neat-HDPE

5% CaCO3 -HDPE

10% CaCO3 -HDPE

20% CaCO3 -HDPE

Heat of fusion (J/g) Percent crystallinity DSC (XRD)* Crystallization temperature (◦ C) Melting temperature (◦ C) Lamellar thickness (nm) Long range order (nm) Average spherulite size (␮m) Nucleating efficiency (%) Modulus at 50 mm/min (MPa) Yield stress at 50 mm/min (MPa) Percent strain-to-fracture (50 mm/min)/(125 mm/min) Room temperature Izod impact strength (kJ/m2 )

125 43 (46)* 113.9 133.3 8.9 20.6 415 – 1050 24 >8/2.4 3.71

156 53 (56)* 116.8 137.4 10.63 20.1 82 44 1190 26 >8/1.5 4.58

166 57 (59)* 116.7 136.1 11.7 20.5 75 46 1365 27 >8/1.1 4.78

197 67 (66)* 115.5 137.9 13.84 20.7 37 52 1695 24 >8/1.08 5.62

212

M. Tanniru et al. / Materials Science and Engineering A 404 (2005) 208–220

Fig. 1. Nature of spherulites in (a) neat and (b) 20% calcium carbonate-reinforced polyethylene composite.

The above observations imply that significant alignment of particles did not occur in the flow direction of the injectionmolded sample, consistent with their uniform dispersion (see below). An increased ratio of I(104) /I(202) would indicate higher alignment of filler particles, in the flow direction of the injection molded specimens. 3.1.2. Dispersion of calcium carbonate The dispersion of particles in 20% CaCO3 –polyethylene composite is presented in Fig. 2. The particles (indicated with arrows) are uniformly distributed in the polyethylene matrix and do not give an indication of aggregation. It seems from this study that the threshold of aggregation has not been reached and is certainly beyond 20%. Uniform dispersion is important because in case of a matrix with aggregates of particles, the stress field will be concentrated around any aggregates, such that cracks will propagate easily and rapidly, causing premature failure.

3.1.3. Tensile and impact properties The tensile modulus and yield stress data are listed in Table 1. The elastic modulus increased from 1050 MPa in neat polyethylene to 1695 MPa in 20% CaCO3 composite. However, there were insignificant differences in yield strength on reinforcement of polyethylene with calcium carbonate. The minor differences are within the experimental range. The values reported in Table 1 are average values of at least three tests for each condition. In general, the deviation from the mean value was not more than ±10%. We believe that the yield stress remained unaffected on the addition of CaCO3 in neat polyethylene. A similar behavior was observed for wollastonite-reinforced [16–18,22,24,31–34] and talcreinforced [31–34] polypropylenes and ethylene–propylene copolymers [31]. This is discussed below. Earlier work [40] indicated that an increase in crystallinity or increase in spherulite size increases the modulus because large spherulites are considered to have a significantly higher

M. Tanniru et al. / Materials Science and Engineering A 404 (2005) 208–220

213

Fig. 2. Low (a) and high (b) scanning electron micrographs showing distribution of calcium carbonate in 20% calcium carbonate-reinforced polyethylene composite.

load-bearing capability. However, in the present study, there is increase in crystallinity but the spherulite size decreases on reinforcement of polyethylene with calcium carbonate. To rationalize these observations, it is reasonable to say that two mutually opposing forces influence the properties—the reinforcement influence of calcium carbonate and the nucleating effect. The reinforcement has a positive effect on modulus and yield stress, while the nucleating effect of CaCO3 particles has a negative effect. The reinforcement effect easily overwhelms any effect due to smaller spherulite size induced by the CaCO3 . The crystallinity increases even though the spherulite size becomes smaller because of the significantly higher nucleation density induced by the CaCO3 mineral particles.

Friedrich [41] first emphasized the effect of morphology and provided strong evidence that semi-crystalline polymers with small spherulites tend to be tougher than those with coarse spherulites because larger spherulites have weak boundaries. While Ouderni and Philips [42] studied the effect of crystallinity and spherulite size individually in polypropylene and observed that an increase in crystallinity or spherulite size decreased the toughness, consistent with Friedrich’s conclusion [41]. The above observations of Friedrich [41] and Ouderni and Philips [42] seems to be applicable for neat polymers but not for the composites. In our case, there is a decrease in spherulite size and increase in crystallinity both of which according to the observations of Friedrich [41] and Ouderni

214

M. Tanniru et al. / Materials Science and Engineering A 404 (2005) 208–220

and Philips [42] should have mutually opposite effects as discussed above. The observation suggests that the behavior of the composite is not a simple function of spherulite size and crystallinity, but is a complex function of other factors and includes lamellar thickness and crystalline long period. Table 1 shows that the lamellar thickness increases with percentage CaCO3 reinforcement. The lamellar thickness is an important controlling parameter in the activation of yield, and yield stress in neat semi-crystalline polymers is proportional to lamellar thickness [43]. This further supports the viewpoint of mutually opposing effect of reinforcement and nucleating capability of CaCO3 . The reinforcement effect of CaCO3 with consequent increase in percentage crystallinity increases the modulus of the composite; while the nucleating effect of CaCO3 decreases the spherulite size, which has a negative effect on yield stress and neutralizes the positive influence of increase in percentage crystallinity on yield stress.

Room temperature Izod impact strength for polyethylene and 5–10% CaCO3 -polyethylene is summarized in Table 1. It can be clearly seen that impact strength increases with increase in %CaCO3 in the polyethylene matrix. Such behavior was attributed to particle-induced cavitation and fibrillation in the composite [36]. 3.2. Micromechanism of scratch deformation in neat and calcium carbonate-reinforced polyethylene Representative SEM micrographs of the scratch morphology of neat and 5–20% calcium carbonate-reinforced polyethylene under identical conditions of scratch test are presented in Figs. 3–6. Figs. 3–6 show periodic ripple marks having the resemblance of the parabolic shape with cracks initiating at both the ends of the scratch tracks and extending towards the center of the scratch. The continuity of the parabolic scratch tracks is disrupted or fragmented by a

Fig. 3. Scanning electron micrographs of the scratch deformation region of neat high density polyethylene at a test load of 4 kgf.

M. Tanniru et al. / Materials Science and Engineering A 404 (2005) 208–220

215

Fig. 4. Scanning electron micrographs of the scratch deformation region of 5% CaCO3 -reinforced high density polyethylene at a test load of 4 kgf.

plastic flow process that occurs within the scratch. At higher magnification, these regions contain ridges/wrinkles that are perpendicular to the scratch direction. However, while periodic scratch tracks were observed in all the materials, the effects of calcium carbonate on the scratch marks on the surface can be summarized as follows: (a) the width of the scratch decreases with increase in calcium carbonate; (b) the cracking at the end of parabolic cracks decreases with increase in calcium carbonate; (c) the severity of parabolic scratch track (height of the track and edge cracks) decreases with consequent decrease in the sequential accumulation of ploughed material at each consecutive parabolic track; (d) the decrease in periodic–parabolic scratch track process with increase in %CaCO3 is accompanied by plastic flow

or tearing of material, referred as fibrillation (see below), that involves lower volume of the material and shallow scratch (Table 3). This plastic flow process decreases the scratch roughness. The above observations clearly suggest that reinforcement of high density polyethylene with calcium carbonate has a strong influence on scratch deformation process. The periodic parabolic scratch tracks can be explained in terms of stick-slip motion between the tip of the indenter and the surface of the material. In the stick-slip process [11], during the stick stage there is no relative motion between the indenter tip and the surface of the material. But the indenter continues to apply stress on the sample surface resulting in deformation of the material underneath the indenter. The tangential or horizontal stress acting during the stick stage is less than the critical stress, but increases with time. Once the stress

216

M. Tanniru et al. / Materials Science and Engineering A 404 (2005) 208–220

Fig. 5. Scanning electron micrographs of the scratch deformation region of 10% CaCO3 -reinforced high density polyethylene at a test load of 4 kgf.

applied by the indenter on the sample surface exceeds the required critical stress, the slip stage initiates resulting in relative motion between the tip of the indenter and the sample surface. The slip stage terminates once the applied stress falls below the critical stress resulting in stick stage when the indenter and material surface stick again. The material gets piled-up in front of the indenter during the slip stage. Thus, the periodic scratch tracks reveal the sequential accumulation and release of tangential force as depicted in Fig. 7. It is clear from the SEM micrographs that the addition of calcium carbonate to high density polyethylene encourages continuous plastic flow of the material over the discontinuous or quasi-static sequential accumulation of material. This behavior is attributed to increase in the toughness and microplasticity of calcium carbonate-reinforced polyethylene composites (Table 1), where the toughening is related to small spherulite size (Fig. 1), and is discussed elsewhere [36].

The increased toughness of calcium carbonate-reinforced polyethylene triggers tearing of matrix ligaments resulting in stretching of fibrils (fibrillation) such that in 20% CaCO3 –polyethylene composite (Fig. 6), fibrillation is the dominant mode of scratch deformation process. Additionally, particle induced cavitation or microvoids contribute to energy absorption and microplastcity. The behavior during scratch tests is consistent with tensile loading. The mode of deformation in neat polyethylene was primarily crazing-tearing while in calcium carbonate containing polyethylene, it was fibrillation [44]. The formation of wrinkle-type ridges in the fibrillation region observed at higher magnification represent stress relaxation (or stress release) [28] of the matrix. The density of these ridges decreased with increase in percentage calcium carbonate because of the increased stiffness or modulus of the matrix (Table 1) that minimizes the build up of stress.

M. Tanniru et al. / Materials Science and Engineering A 404 (2005) 208–220

217

Fig. 6. Scanning electron micrographs of the scratch deformation region of 20% CaCO3 -reinforced high density polyethylene at a test load of 4 kgf.

It is discussed by Hamilton and Goodman [45] and Xiang et al. [3] that along the scratch length, a significantly higher tensile stress is generated at the tail end of the scratch (behind the indenter) during scratch deformation than that caused by the indentation. The magnitude of the maximum tensile stress is higher for high modulus polymers and is applicable to

calcium carbonate-reinforced polyethylene. During scratch deformation, when the tensile stress behind the scratch is greater than the tensile strength of the polymeric material, plastic deformation occurs. From the modulus considerations, the surface of calcium carbonate-reinforced polyethylene composite experiences higher tensile stress because of

Table 3 Comparison of scratch deformation parameters of neat and CaCO3 -reinforced polyethylene composites Scratch deformation parameters

Maximum depth of the scratch (␮m) Average width of the scratch (mm) RMS scratch roughness from AFM (nm) Average scratch roughness from AFM (nm) Percent elastic recovery from Eq. (6) Scratch hardness from Eq. (7) (kgf/mm2 )

Neat-HDPE

5% CaCO3 -HDPE

10% CaCO3 -HDPE

20% CaCO3 -HDPE

2 kgf

4 kgf

2 kgf

4 kgf

2 kgf

4 kgf

2 kgf

4 kgf

83 1.56 354 294 5.9 1.11

191 1.65 581 425 4.9 0.98

75 1.37 282 228 6.7 1.23

149 1.42 522 407 6.6 1.25

56 1.26 277 214 8.1 1.64

106 1.32 451 392 8.5 1.74

43 1.23 225 156 9.0 2.13

79 1.24 377 278 9.6 2.33

218

M. Tanniru et al. / Materials Science and Engineering A 404 (2005) 208–220

Fig. 7. Schematic depiction of the periodic scratching process.

the increased stiffness. While the transformation of parabolic scratch tracks in neat high density polyethylene to fibrillation in calcium carbonate-reinforced polyethylene can be attributed to enhancement of toughness; the lower depth of the scratch (Table 2), reduction in cracking at the scratch edge of scratch tracks, and lower density of ridges, suggests the favorable effect of higher modulus in polymer composite materials. This can be related to shift in the von Mises stress from the surface to subsurface resulting in a reduction in the maximum tensile stress induced by the scratch and consequently reduced surface damage. Another possibility is that the higher toughness of calcium carbonate-reinforced polyethylene composites reduces or dissipates the tensile stress that is generated at the tail end of the scratch (Fig. 7). Transmission electron microscopy observations of the near surface microstructure in scratched PP blends have indicated the extension of deformation fields into the bulk and are characterized by dilation of material inside the shear bands [46]. The extensive cracking at the ends of the parabolic scratch tracks in neat polyethylene also suggests a higher level of tensile stress on the surface during scratching whose magnitude is either reduced or shifts to sub-surface region on reinforcement of polyethylene with CaCO3 . 3.3. Evaluation of scratch damage Scratch damage can be quantified in terms of percentage elastic recovery and scratch hardness using the maximum depth of the scratch and width of the scratch. Additionally, average surface roughness is also a measure of the magnitude of surface deformation. The width and maximum depth of the scratch was determined by SEM and surface roughness by atomic force microscopy using a field of view of 30 ␮m × 30 ␮m along the length of the scratch. The maximum depth was measured by examining the sample in crosssection. From the quantitative determination of the above parameters, the scratch resistance of the material can be predicted. It may be noted from the data presented in Table 2. The decreasing values of average scratch roughness, depth of the scratch, and scratch width imply that susceptibility to scratch deformation decreases with increase in percentage reinforcement. The lower depth of the scratch is also consistent with

greater elastic recovery (Table 2). The elastic recovery, he , was calculated using the equation of linear elastic theory, and is given by Eq. (6): he =

Fn (1 − ν2 ) 2EW

(6)

where Fn is the normal applied load, ν is Poisson’s ratio, E is modulus of elasticity and W is the scratch width. Thus, from the observation presented above in Figs. 3–6 one can predict the resistance to mechanically induced surface deformation. It may, however be noted that since the contact area experiences visco-elastic plasticity, the part that is elastic–plastic recovers partly instantly and partly after some time delay. The elastic part that is recovered is directly related to the contact width of the indenter and depends on the scratch velocity. Scratch hardness can also be used to predict the scratch resistance of the material and is defined by: H=


Fn (2rd − d 2 )

(7)

where Fn is the scratch load in kgf, d is the depth of the scratch in mm, and r is the radius of the stylus in mm for the Hoffman stylus. Eq. (7) involving depth of the scratch is sensitive to the nature of the material [35]. The scratch hardness predicted using Eq. (7) is listed for all the four materials is presented in Table 2. There is a direct relationship between scratch depth and scratch hardness (Table 2). Materials with higher scratch hardness are expected to exhibit greater resistance to scratch damage and hence exhibit lower scratch visibility. The values of scratch parameters from Table 2 suggest that the extent of improvement was maximum for 20% CaCO3 -reinforced high density polyethylene. 3.4. Quantification of stress whitening during scratch damage We have adopted our recently developed approach involving optical imaging and gray level determination to quantify the stress whitening. Quantitative plots comparing the change in gray level of scratched sample with respect to unscratched specimen is presented in Fig. 8 for 2 and 4 kg loads. The

M. Tanniru et al. / Materials Science and Engineering A 404 (2005) 208–220

219

Fig. 8. Change in gray level vs. distance from scratch deformation region of neat high density polyethylene and calcium carbonate-reinforced high density polyethylene composites under identical test conditions.

relative change in gray level decreased with increase in percentage CaCO3 suggesting that CaCO3 -reinforced HDPE exhibits greater resistance to stress whitening (minimum scratch visibility) consistent with SEM and AFM observations. It may however, be noted that the scratch visibility depends on the deformation features that can scatter light. A material with lower scratch resistance need not necessarily exhibit reduced scratch visibility. In our case, higher scratch resistance is directly related to lower scratch visibility. The micromechanism of stress whitening involves combined light scattering effects from deformed features such as microcracks, microvoids, and stretched fibrils. The relationship between light scattering and deformation is, however, complex since it may be influenced by the size of the microvoids, and also by the change in the refractive index induced by molecular level changes that concern entanglement of polymer chains and reorientation of polymer molecules, or difference in refractive index between matrix and reinforcement minerals. Deformation features of size similar to the wavelength of visible light or entities that have refractive index different from that of air and are of size corresponding to wavelength of light will promote scattering of visible light giving the appearance of a whitened zone. Surface roughness is another factor that can contribute to light scattering. It is clear from Figs. 3–6 that the transformation from predominantly quasi-periodic scratch tracks with cracking at the edges of the scratch in neat polyethylene to continuous plastic flow in 20% CaCO3 –polyethylene composites decreases the average scratch roughness (Table 2). This transformation can be visualized as “ironing” effect (ironing of parabolic scratch tracks) induced by the CaCO3 that reduces the surface roughness. Also, in general small ridges observed at high magnification perpendicular to the scratch direction tend to impart the sample surface ‘rougher’ than the broken fibrils, as since

shown recently by profilometry studies [47]. Since the size or thickness of these ridges is similar to the wavelength of light, they are expected to encourage and enhance the scattering process. The high scratch visibility of neat high density polyethylene can be attributed to the high density of ridges, microcracks at the end of parabolic scratch tracks, and higher average surface roughness.

4. Conclusions 1. The reinforcement of high density polyethylene with 5–20% calcium carbonate has an overriding influence on tensile modulus of polymer composites that increases percentage crystallinity, elastic recovery, and increases resistance to scratch deformation. 2. The scratch morphology of neat high density polyethylene is characterized by periodic ripple type scratch tracks with cracks at the ends of the scratch tracks and extending towards the center. The periodic nature of the scratch tracks can be explained in terms of stick-slip motion between the tip of the indenter and the material surface resulting in sequential accumulation and release of tangential force. 3. The reinforcement of polyethylene with calcium carbonate disrupts the continuity of the ripple-type scratch track by encouraging fibrillation that involves plastic deformation of a lower volume of the material and consequently shallow scratch. 4. A comparative assessment of scratch damage in terms of maximum depth of the scratch, average scratch roughness, and scratch hardness suggests increase in resistance to scratch deformation of calcium carbonate-reinforced polyethylene composite. The scratch hardness is a

220

M. Tanniru et al. / Materials Science and Engineering A 404 (2005) 208–220

relevant parameter that can be used to determine resistance to scratch deformation. 5. Microvoids and other surface features such as wrinkles/ridges are the primary source of light scattering resulting in stress whitening.

References [1] B.J. Briscoe, P.D. Evans, E. Pelilo, S.K. Sinha, Wear 200 (1996) 137. [2] J. Chu, C. Xiang, H.J. Sue, D. Hollis, Polym. Eng. Sci. 40 (2000) 944. [3] C. Xiang, H.J. Sue, J. Chu, K. Masuda, Polym. Eng. Sci. 41 (2001) 23. [4] A. Dasari, J. Rohrmann, R.D.K. Misra, Macromol. Mater. Eng. 287 (2002) 889. [5] A. Dasari, S.J. Duncan, R.D.K. Misra, Mater. Sci. Technol. 18 (2002) 1227. [6] C. Gauthier, R. Schirrer, J. Mater. Sci. 35 (2000) 2121. [7] A. Dasari, J. Rohrmann, R.D.K. Misra, Macromol. Mater. Eng. 289 (2002) 887. [8] A. Dasari, S.J. Duncan, R.D.K. Misra, Mater. Sci. Technol. 19 (2003) 239. [9] A. Dasari, J. Rohrmann, R.D.K. Misra, Mater. Sci. Eng. A 354 (2003) 167. [10] P.Z. Wang, Ph.D. thesis, University of Cambridge, 2002. [11] K. Li, B.Y. Ni, J.C.M. Li, J. Mater. Res. 11 (1996) 1574. [12] S. Jacobson, M. Olsson, P. Hedenquist, O. Vingsbo, ASM Handbook, vol. 18, ASM International, Materials Park, OH, 1992, p. 430. [13] B.J. Briscoe, E. Pelillo, F. Ragazzi, S.K. Sinha, Polymer 39 (1998) 2161. [14] R.J. Young, P.A. Lowell, Introduction to Polymers, second ed., Chapman & Hall, 1991, p. 177. [15] A. Dasari, J. Rohrmann, R.D.K. Misra, Mater. Sci. Eng. A 351 (2003) 200. [16] A. Dasari, J. Rohrmann, R.D.K. Misra, Mater. Sci. Eng. A 358 (2003) 372. [17] A. Dasari, J. Rohrmann, R.D.K. Misra, Mater. Sci. Eng. A 358 (2003) 356. [18] R.S. Kody, D.C. Martin, Polym. Eng. Sci. 36 (1996) 298. [19] C. Xiang, H.-J. Sue, J. Chu, B. Coleman, J. Polym. Sci. B: Polym. Phys. 39 (2001) 47.

[20] P. Rangarajan, K. Harding, V. Watkins, APS Bull. 46 (2001) 157. [21] P.Z. Wang, I.M. Hutchings, S.J. Duncan, L. Jenkins, E. Woo, Proceedings of the SPE Automotive TPO Global Conference, 2000, p. 107. [22] P.Z. Wang, I.M. Hutchings, S.J. Duncan, L. Jenkins, Proceedings of the SPE Automotive TPO Global Conference, Detroit, USA, 2001, p. 107. [23] A. Dasari, J. Rohrmann, R.D.K. Misra, Mater. Sci. Eng. A 360 (2003) 237. [24] X.D. Zhang, G. Shi, Polymer 35 (1994) 5067. [25] C.B. Ng, L.S. Schadler, R.W. Siegel, Adv. Comp. Lett. 10 (2001) 101. [26] Z. Gao, H. Tsou, J. Polym. Sci. B: Polym. Phys. 37 (1999) 155. [27] A. Dasari, J. Rohrmann, R.D.K. Misra, Mater. Sci. Eng. A 364 (2004) 357. [28] A. Dasari, R.D.K. Misra, Acta Mater. 52 (2004) 1683. [29] H. Nathani, A. Dasari, R.D.K. Misra, Acta Mater. 52 (2004) 3217. [30] R.D.K. Misra, H. Nathani, A. Dasari, S.D. Wanjale, J.P. Jog, Mater. Sci. Eng. A 386 (2004) 175. [31] A. Dasari, J. Rohrmann, R.D.K. Misra, Polym. Eng. Sci. 44 (2004) 1738. [32] R.S. Hadal, A. Dasari, J. Rohrmann, R.D.K. Misra, Mater. Sci. Eng. A 380 (2004) 326. [33] R.S. Hadal, R.D.K. Misra, Mater. Sci. Eng. A 374 (2004) 374. [34] R.S. Hadal, A. Dasari, J. Rohrmann, R.D.K. Misra, Mater. Sci. Eng. A 380 (2004) 326. [35] R.D.K. Misra, H. Nathani, A. Dasari, S.D. Wanjale, J.P. Jog, Mater. Sci. Eng. A 386 (2004) 175. [36] M. Tanniru, R.D.K. Misra, submitted for publication. [37] D.C. Bassett, Principles of Polymer Morphology, Cambridge University Press, Cambridge, UK, 1981. [38] B. Fillon, J.C. Wittmann, B. Lotz, A. Thierry, J. Polym. Sci. B: Polym. Phys. 31 (1993) 1383. [39] B. Fillon, B. Lotz, A. Thierry, J.C. Wittmann, J. Polym. Sci. B: Polym. Phys. 31 (1993) 1395. [40] J.L. Way, J.R. Atkinson, J.J. Nutting, J. Mater. Sci. 9 (1974) 293. [41] K. Friedrich, Adv. Polym. Sci. 52–53 (1983) 225. [42] M. Ouderni, P.J. Philips, J. Eng. Appl. Sci. 2 (1996) 2312. [43] B. Lotz, A.J. Kovacs, J.C. Wittmann, J. Polym. Sci., Polym. Phys. Ed. 909 (1975) 13. [44] R.D.K. Misra, P. Nerikar, K. Bertrand, D. Murphy, Mater. Sci. Eng. A 384 (2004) 284. [45] G.M. Hamilton, L.E. Goodman, J. Appl. Mech. 33 (1966) 371. [46] H. Tang, D.C. Martin, J. Mater. Sci. 38 (2003) 803. [47] P.Z. Wang, Ph.D. thesis, University of Cambridge, 2002.