Acta Polytechnica CTU Proceedings doi:10.14311/APP.2018.18.0066 Acta Polytechnica CTU Proceedings 18:66–71, 2018 © Czech Technical University in Prague, 2018 available online at http://ojs.cvut.cz/ojs/index.php/app EXPERIMENTAL INVESTIGATION OF THE FAILURE BEHAVIOUR OF POLYPROPYLENE COMPOUNDS FOR INSTRUMENTED PUNCTURE TESTS Florian Kiehasa, ∗, Anna Kalteisa, Michael Jerabekb, Zoltán Majora a Institute for Polymer Product Engineering, Altenberger Straße 69, 4040 Linz, Austria b Borealis Polyolefine GmbH, Sankt-Peter-Straße 25, 4020 Linz, Austria ∗ corresponding author: florian.kiehas@jku.at Abstract. Instrumented puncture tests according to ISO 6603-2 and ASTM D3763 were executed for five different Polypropylene compounds (talcum-, glass fibre- and elastomer modified) with specimen thicknesses ranging from 1 mm to 4 mm. Over 1500 puncture tests were executed at the Impact & Long-term Behaviour laboratory of the company Borealisr in Linz. This serves as strong foundation for statistical evaluations of the ductile/brittle transition temperature. For different materials and ductile/brittle transition determination methods, similar trends have been observed, which were characterized by introducing shift factors. Keywords: polypropylene, compound, puncture, mechanical testing, ductile/brittle transition. 1. Introduction For polymers, the energy absorption capacity strongly depends on the morphology of the plastic and the experimental environment. In the case of impact or puncture tests, temperature and strain rate variations have a significant influence on the failure behaviour: Almost all unoriented plastics exhibit brittle behaviour at low temperatures as well as at high strain rates [1]. Mechanical measurements of different puncture test standards result in distinct force-deflection curves, which in turn give different transitions from ductile to brittle. Furthermore, there is no concise method for the calculation of an exact transition temperature. It is assumed to be somewhere in the transition regime, a temperature range where sudden drops in energy absorption capacity indicate a change of impact be- haviour from ductile to brittle. First, a precise method for the calculation has to be introduced. Then, the difference of the transition as a function of specimen thickness, temperature, test standard and other pa- rameters can be determined. 2. Experimental Investigation 2.1. Test Parameters The test setup and parameters of ISO 6603-2 and ASTM D3763 vary greatly. The main difference is the striker- and support ring geometry. For this refer- ence analysis, tests for both standards were executed clamped, lubricated and at a speed of 4400 mm/s to allow for comparison. Therefore, the ASTM D3763 was slightly modified. A summary of all parameters can be looked up in Table 1 and Figure 1 shows the different geometries of each test standard. A more de- tailed description of each test standard can be looked up in literature [2, 3]. ISO modified* ASTM Specimen dimensions 60×60 mm ∅102 mm Specimen thickness 1,2,3 mm 1,2,3,4 mm* Striker diameter 20.0 mm 12.7 mm Support ring diameter 40.0 mm 76.0 mm Test speed 4400 mm/s 4400 mm/s* Clamped yes yes Lubricated yes yes* Table 1. Selected test conditions and parameters. Figure 1. Striker and support ring geometries ISO (left) and ASTM (right). 2.2. Materials and test schedule An extensive test program with a total number of 1586 specimens was carried out. The work- load was divided between two machines: a servohy- draulic ROELL/AMSLER High Speed and a gravi- tative (spring driven) CEAST Fractovis Plus. The comparability of measurement data obtained by both machines is documented by [4]. Five different polypropylene grades developed and produced by the company Borealisr were selected for this analysis. Three materials are in-reactor made het- erophasic copolymers, which are different in rubber- and matrix design (elastomer modified polypropylene M1 to M3). M5 is also an in-reactor made heteropha- sic copolymer filled with talc. M4 is a glass filled polypropylene. 66 http://dx.doi.org/10.14311/APP.2018.18.0066 http://ojs.cvut.cz/ojs/index.php/app vol. 18/2018 Investigation of PP-compounds for Instrumented Puncture Tests Four specimen thicknesses from 1 mm to 4 mm were tested with the exception of 4 mm ISO 6603-2 speci- mens, for which no injection moulding tool was avail- able for manufacturing. 1 mm thick glass-fibre rein- forced ISO plaques were also not available. Examined conditioning temperatures range from -70 ◦C to 23 ◦C. Eight temperature levels with five repetitions respec- tively were executed for each material, test standard and specimen thickness. In certain cases, additional measurements with three repetitions were carried out in the transition regime. Altogether, there are 34 test series with descriptions following Table 2: Polypropylene compound Test standard Thickness M1(Elastomer modified PP) ISO(ISO 6603-2) t1(1mm) M2(Elastomer modified PP) ASTM(ASTMD3763) t2(2mm) M3(Elastomer modified PP) t3(3mm) M4(Glass fibre reinforced PP) t4(4mm) M5(Talcum reinforced PP) Table 2. Nomenclature of test series. For example, M4ASTMt3 represents a test series of material M4 with 3 mm thick specimens tested under the modified ASTM D3763 standard conditions. 2.3. Data evaluation For both test standards, the failure criterion is 50 % force drop after peak load Fm. The corresponding puncture energy Ep is obtained by integrating the force up to the deflection at puncture sp. The transition from elastic to elastic-plastic mate- rial behaviour was described with the introduction of a yield equivalent point (sy/Fy)[5, 6]. The initial slope klin (corresponding to the relative stiffness [3]) of the curve can be approximated through the specifi- cation of another characteristic data point (sy0/Fy0), which is necessary to avoid deviations caused by dis- continuities at the start of impact. These points are defined as having the largest orthogonal distance from the direct connection line between origin (0/0) and peak load (sm/Fm). Thus, the energy Em expended up to the maximum load can be split into an elastic part Eel and a plastic part Epl. The graphical determination of the starting and end point of the initial slope, as well as the geo- metric interpretation of the plastic and elastic energy contributions are illustrated in Figure 2. Following characteristic values are examined: • Maximum force/deflection/energy Fm/sm/Em • Puncture force/deflection/energy Fp/sp/Ep • Initial force/deflection Fy0/sy0 • Yield equivalent force/deflection Fy/sy • Elastic/inelastic energy portions Eel/Epl • Relative stiffness klin • Width of plastic plateau ∆sw klin = Fy − Fy0 sy − sy0 (1) Eel = F 2m 2klin (2) Epl = Em − Eel (3) Figure 2. Evaluation of force-deflection data. Two characteristic energy ratios have been introduced to describe the elastic properties as well as the failure behaviour of force-deflection data: the elastic energy ratio Rel and the ductility index [7] Rduc are specified in the following equations: Rel = Epl + Eel Eel (4) Rduc = Ep − Em Ep (5) 2.3.1. Curve characterization The ISO 6603-2 curve type evaluation is a common method to characterize the force/deflection data ob- tained by puncture test ecperiments. It provides four typical types of curve progression that can usually be observed during data evaluation. In general, these range from ductile to brittle and focus on information like yielding, crack initiation and crack propagation [2]. Obtained data has been divided into these four categories, which are compiled in Figure 3. The determination of the ductile/brittle transition by means of curve type characterization is described in section 2.4.5. 2.3.2. Optical characterization The different types of fracture appearance are taken from a company-internal standard developed by Gen- eral Motors Corporation. It grades fracture appear- ances from 1 to 11 and covers almost all kinds of break that may occur during impact testing. A subdivision ranging from ductile, to semi-ductile, up to brittle can be made. Illustrations of the individual fracture types are depicted in Figure 4. The optical characterization was utilized to deter- mine the ductile/brittle transition as well, which is described in section 2.4.6. 67 F. Kiehas, A. Kalteis, M. Jerabek, Z. Major Acta Polytechnica CTU Proceedings Figure 3. Curve types: YD Yielding with deep draw- ing (top left), YS Yielding with stable crack (top right), YU Yielding with unstable crack (bottom left), NY No yielding (bottom right) [2]. 1 2 3 4 5 6 7 8 9 10 11 Figure 4. Fracture appearances: from ductile (1 - 5), to ductile-fragile (6 - 7), to fragile-ductile (8 - 9), up to fragile (10 - 11). 2.4. Ductile/brittle transition temperature Various methods for the determination of the duc- tile/brittle transition temperature have been proposed in literature. For impact tests with steep inclines in the transition regime, not only the tangent formation of the upper- and lower shelf impact strengths [8, 9] but also optical investigations of the fracture appear- ance was investigated [10]. In [11] the inflection point of the temperature-dependent course was determined. Likewise, the penetration energy was proposed for exploration. For this reference analysis, six different methods for the determination of the ductile/brittle transition tem- perature have been introduced and will be listed in the following subsections. It should be pointed out, that values are interpolated between data points. Figure 5 depicts graphic representations for the determination of each transition temperature. 2.4.1. Puncture energy The arithmetic mean of the local minimum and maxi- mum puncture energy Ep is calculated and the transi- tion is assumed at the corresponding condition tem- Figure 5. Determination of transition temperatures. perature T E. 2.4.2. Elastic energy ratio The peak load energy Em is compared to the elas- tic energy Eel. It should be pointed out that the minimum value for this energy ratio is 1 when the plastic portion disappears. This case corresponds to absolutely no yielding before peak load with the cor- responding curve type NY No yielding. Threshold values of 1.4, 1.2 and 1.0 were examined, with T Rat1.2 giving the most consistent results. 2.4.3. Ductility index In contrast to the elastic energy ratio Rel, the ductil- ity index Rduc is only obtained from information of the force deflection data after peak load. Therefore, yielding before peak load can not be detected with this method and is neglected. A ductility index of 0 implies, that the peak load is directly followed by an unstable crack, and consequently brittle fracture. Threshold values of 40 %, 30 % and 5 % were examined, with T Duc30 giving the most consistent results. 2.4.4. Deflection For completely brittle fractures, all characteristic events of a puncture test happen in a narrow time frame. The deflection difference between sm, sp and stotal (as well as sy when there is no plastic plateau) becomes very small. T s is the temperature where sm and sp converge. 2.4.5. Curve characterization This method for the ductile/brittle transition tem- perature determination takes advantage of the elabo- rate ISO6603-2 curve characterization introduced in section 2.3.1. The curve types usually occur in the sequence of YD, YS, YU and NY from high to low 68 vol. 18/2018 Investigation of PP-compounds for Instrumented Puncture Tests temperatures. An attempt was made to describe the ductile/brittle behaviour by looking out for the tran- sitions from YU Yielding with unstable crack (bottom left) to NY No yielding. This leads to two specific temperatures TY U and TN Y , which are triggered at the first appearance of their respective curve type. 2.4.6. Optical characterization TGM corresponds to the test temperature, where a fragile fracture type occurs for the first time. 2.5. Overview transition temperatures A compilation of all calculated transition tempera- tures is given in Table 3. With the exception of the glass fibre reinforced material M4, all materials follow similar trends between test standards and specimen thicknesses. Reproducibility, consistency and compa- rability are the main factors that need to be considered when evaluating the transition temperatures. At the same time, the calculations involved should be as simple as possible. T Duc30 shows the most inconsistent results because it is sensitive to noise of the force-deflection data after crack initiation. Despite occasional outliers, the ISO characterization has proven itself to be an adequate representation of the transition regime with most of the other transition temperatures lying between the boundary values of TY U and TN Y . For thin specimens, TN Y and TY U stretch over a wide temperature range, which gets more narrow for thicker specimens. In comparison to all the other categories, the frac- ture appearance usually gives much higher transition temperatures TGM , because fractures are classified as brittle as soon as crack initiation is observed disre- garding any plasticity that might be present before. 2.6. Transition shift factors Since the trends of all categories are strikingly sim- ilar as can be seen in Table 3, the shift factors are contemplated to be universally applicable for all tran- sition temperatures. However, the impact behaviour of glass fibre- and talcum reinforced PP-compounds is vastly different from the other tested materials. Con- sequently, only elastomer modified PP-grades (M1, M2, M3) as well as the most consistent transition temperatures (T E, T Rat1.2 , T s) are used to determine the shift factors. For data extrapolation, 1 mm ISO test series are chosen as initial reference values, since 1 mm test series show the least standard deviations and ISO is generally more consistent than ASTM. This procedure is illustrated in Table 4. Finally, these shift factors are summarized in average values α, β, γ, δ, � and ζ, which are listed in Table 5. It should be pointed out, that the transition from ductile to brittle was not fully covered for the material M2, even at -70 ◦C, which is the reason why T s values for M2ISOt2 and M2ISOt3 are assumed to be -70 ◦C. T E T Rat1.2 T s T Duc30 TN Y TY U TGM M1ISOt1 -47 -50 -49 -35 -52 -30 -30 M1ISOt1 -57 -59 -57 -53 -58 -52 -50 M1ISOt3 -56 -56 -58 -55 -60 -56 -56 M1ASTMt1 -40 -29 -30 18 -40 -23 -30 M1ASTMt2 -47 -45 -50 -41 -50 -40 50 M1ASTMt3 -43 -43 -50 -36 -50 -50 -40 M1ASTMt4 -45 -45 -50 -41 -50 -50 -50 M2ISOt1 -50 -55 -50 -40 - 0 -20 M2ISOt2 -63 -66 -70 -56 - -30 -30 M2ISOt3 -64 -65 -70 - - -40 -30 M2ASTMt1 -40 -36 -37 7 - 23 -20 M2ASTMt2 -55 -55 -60 -41 - -40 -20 M2ASTMt3 -52 -50 -60 -43 - -50 M2ASTMt4 -54 - -60 -22 - -30 -30 M3ISOt1 -19 -20 -17 6 -30 -10 -20 M3ISOt2 -28 -29 -30 -8 -40 -10 -30 M3ISOt3 -26 -27 -25 -17 -34 -20 -30 M3ASTMt1 -23 -3 -10 16 -20 23 -10 M3ASTMt2 -26 -14 -19 5 -20 0 -10 M3ASTMt3 -26 -13 -19 7 -20 -10 -10 M3ASTMt4 -24 - -28 -6 -30 -10 -20 M4ISOt2 -18 - - - - - - M4ISOt3 -7 - - -7 - - - M4ASTMt1 -17 - -17 -11 - - - M4ASTMt2 -7 -27 -52 -27 - - - M4ASTMt3 -6 -14 -5 -1 - - - M4ASTMt4 5 -59 -22 6 - - - M5ISOt1 -35 -35 -27 -4 -36 -10 -30 M5ISOt2 -36 -36 -36 -23 -38 -30 -40 M5ISOt3 -39 - -36 -21 -50 -20 -40 M5ASTMt1 -35 -25 -21 11 -30 23 -30 M5ASTMt2 -35 -34 -30 -17 -40 -20 -30 M5ASTMt3 -34 -30 -39 -14 -40 -20 -30 M5ASTMt4 -27 - -39 -12 -40 -20 -20 Table 3. Overview of all transition temperatures (all values in Celsius); nomenclature e.g. M3ISOt1: material M3, ISO test standard and 1 mm thick speci- mens. T E T Rat1.2 T s M1ISOt1 226 = x 223 = x 224 = x M1ISOt1 216 = x · 0.955 214 = x · 0.959 216 = x · 0.964 M1ISOt3 217 = x · 0.960 217 = x · 0.973 215 = x · 0.959 M1ASTMt1 233 = x · 1.030 244 = x · 1.094 243 = x · 1.084 M1ASTMt2 226 = x · 1.000 228 = x · 1.022 223 = x · 0.995 M1ASTMt3 230 = x · 1.017 230 = x · 1.031 223 = x · 0.995 M1ASTMt4 228 = x · 1.008 228 = x · 1.022 223 = x · 0.995 M2ISOt1 223 = x 218 = x 223 = x M2ISOt2 210 = x · 0.941 207 = x · 0.949 208 = x · 0.932 M2ISOt3 209 = x · 0.937 207 = x · 0.954 208 = x · 0.932 M2ASTMt1 233 = x · 1.044 237 = x · 1.087 236 = x · 1.058 M2ASTMt2 218 = x · 0.977 218 = x · 1.000 213 = x · 0.955 M2ASTMt3 221 = x · 0.991 223 = x · 1.022 213 = x · 0.955 M2ASTMt4 219 = x · 0.982 213 = x · 0.955 M3ISOt1 254 = x 253 = x 256 = x M3ISOt2 245 = x · 0.964 244 = x · 0.964 243 = x · 0.949 M3ISOt3 247 = x · 0.972 246 = x · 0.972 248 = x · 0.968 M3ASTMt1 250 = x · 0.984 270 = x · 1.067 263 = x · 1.027 M3ASTMt2 247 = x · 0.972 259 = x · 1.023 254 = x · 0.992 M3ASTMt3 247 = x · 0.972 260 = x · 1.027 254 = x · 0.992 M3ASTMt4 249 = x · 0.980 245 = x · 0.957 Table 4. Determination of individual shift factors (all values in Kelvin). avg. std. α 0.953 0.011 ISO1mm to ISO2mm β 0.958 0.015 ISO1mm to ISO3mm γ 1.053 0.035 ISO1mm to ASTM1mm δ 0.993 0.022 ISO1mm to ASTM2mm � 1.000 0.026 ISO1mm to ASTM3mm ζ 1.005 0.038 ISO1mm to ASTM4mm Table 5. Average Kelvin temperature shift factors α, β, γ, δ, � and ζ and their standard deviations. The standard deviation of the shift factors listed in Table 5 become greater for thicker specimens. The predictions from 1 mm ISO to thicker ISO series shows accurate predictions, while extrapolations from ISO to ASTM generally hold higher deviations. Figure 6 compares the predictions of the average shift factors with the results of the measured data. 69 F. Kiehas, A. Kalteis, M. Jerabek, Z. Major Acta Polytechnica CTU Proceedings (a) M1 (b) M2 (c) M3 Figure 6. Shift factor predictions versus reality: The trend between ISO 1 mm (black) and thicker specimens (purple) can be predicted (blue) with temperature shift factors α, β, γ, δ, � and ζ. 3. Conclusions The experimental investigations show that transition temperatures for elastomer modified materials hold similar trends between test standards and specimen thicknesses. Temperature shift factors can be estab- lished to allow for the prediction of transition tem- peratures for elastomer filled PP at different testing conditions. While the data extrapolation is more con- sistent for ISO 6603-2 than ASTM D3763, satisfying values for all tested materials and conditions can be achieved. This may drastically reduce testing time, effort and costs and allow for comparison of data obtained by different test standards. List of symbols Fm Peak load [kN] sm Deflection at peak load [mm] Em Energy up to peak load [J] Fp Puncture force [kN] sp Puncture deflection [mm] Ep Puncture energy [J] Fy0 Initial force [kN] sy0 Initial deflection [mm] Fy Yield equivalent force [kN] sy Yield equivalent deflection [mm] Eel Elastic energy portion [J] Epl Plastic energy portion [J] klin Relative stiffness [N/mm] ∆sw Width of plastic plateau [mm] Rel Elastic energy ratio [–] Rduc Ductility index [–] stotal Total deflection [mm] Y D Yielding with deep drawing Y S Yielding with stable crack growth Y U Yielding with unstable crack growth NY No yielding T Temperature [◦C] T E Transition temperature; puncture energy [◦C] T Rat1.2 Transition temperature; elastic energy ratio [◦C] T Duc30 Transition temperature; ductility index [◦C] TGM Transition temperature; optical [◦C] TN Y Transition temperature; curve type NY [◦C] TY U Transition temperature; curve type YU [◦C] α,β,γ,δ,�,ζ Average shift factors [–] Acknowledgements This research was supported by Borealis Polyolefine GmbH (Linz, A.) References [1] E. Baur, T. A. Osswald, N. Rudolph. Saechtling Kunststoff Taschenbuch. 31st printing. Carl Hanser Verlag, 2013. [2] Plastics -Determination of puncture impact behaviour of rigid plastics - Part 2: Instrumented impact testing. ISO 6603-2, International Organization for Standardization, Geneva, Switzerland, 2000. [3] Astm d3763-15, standard test method for high speed puncture properties of plastics using load and displacement sensors. ASTM International 2015. doi:10.1520/D3763-15. [4] D. Huber. Temperaturgeführte Durchstoßprüfung von Kunststoffen - Vergleich von Durchstoßprüfmaschinen, Bachelor of Science, Montanuniversität Leoben, 2009. [5] M. Böhning, U. Niebergall, A. Adam, W. Stark. Influence of biodiesel sorption on temperature-dependent impact properties of polyethylene. Polymer Testing 40:133 – 142, 2014. doi:10.1016/j.polymertesting.2014.09.001. [6] V. Hristov, R. Lach, W. Grellmann. Impact fracture behavior of modified polypropylene/wood fiber composites. Polymer Testing 23(5):581 – 589, 2004. doi:10.1016/j.polymertesting.2003.10.011. [7] A. Pavan, J. Williams. Fracture of Polymers, Composites and Adhesives. Volume 27. Elsevier Science, 2000. [8] A. van der Wal, A. Wal, J. Mulder, R. Gaymans. Fracture of polypropylene: 2. the effect of the crystallinity. Polymer 39(22):5477–, 1998. doi:10.1016/S0032-3861(97)10279-8. [9] A. Wal, A. van der Wal, J. Mulder, et al. Polypropylene- rubber blends:1. the effect of matrix properties on the impact behaviour. Polymer 39(26):6781–, 1998. doi:10.1016/S0032-3861(98)00170-0. [10] A. van der Wal, A. Wal, R. Gaymans. Polypropylene-rubber blends 5: deformation mechanism during fracture. Polymer 1999(40):6067–6067, 1999. doi:10.1016/S0032-3861(99)00216-5. 70 http://dx.doi.org/10.1520/D3763-15 http://dx.doi.org/10.1016/j.polymertesting.2014.09.001 http://dx.doi.org/10.1016/j.polymertesting.2003.10.011 http://dx.doi.org/10.1016/S0032-3861(97)10279-8 http://dx.doi.org/10.1016/S0032-3861(98)00170-0 http://dx.doi.org/10.1016/S0032-3861(99)00216-5 vol. 18/2018 Investigation of PP-compounds for Instrumented Puncture Tests [11] C. Grein, K. Bernreitner, A. Hauer, et al. Impact modified isotatic polypropylene with controlled rubber intrinsic viscosities: Some new aspects about morphology and fracture. Journal of Applied Polymer Science 87(10):1702–1712. https://onlinelibrary. wiley.com/doi/pdf/10.1002/app.11696 doi:10.1002/app.11696. 71 https://onlinelibrary.wiley.com/doi/pdf/10.1002/app.11696 https://onlinelibrary.wiley.com/doi/pdf/10.1002/app.11696 http://dx.doi.org/10.1002/app.11696 Acta Polytechnica CTU Proceedings 18:66–71, 2018 1 Introduction 2 Experimental Investigation 2.1 Test Parameters 2.2 Materials and test schedule 2.3 Data evaluation 2.3.1 Curve characterization 2.3.2 Optical characterization 2.4 Ductile/brittle transition temperature 2.4.1 Puncture energy 2.4.2 Elastic energy ratio 2.4.3 Ductility index 2.4.4 Deflection 2.4.5 Curve characterization 2.4.6 Optical characterization 2.5 Overview transition temperatures 2.6 Transition shift factors 3 Conclusions List of symbols Acknowledgements References