Printer Friendly

The Brittle-Ductile Transition Temperature of Polycarbonate as a Function of Test Speed.


M. J. J. HAMBERG [**]

The fracture behavior of polycarbonate was studied as a function of temperature (-80[degrees]C to +80[degrees]C) and test speed ([10.sup.-5] to 10 m/s) using an instrumented, singleedged, notched tensile test (SENT). SENT tests give information on the fracture stress, fracture displacements, and fracture energies of polycarbonate, and from these data the average crack speeds were calculated and the brittle-ductile transitions were determined. The fracture stress and the fracture energies of ductile fracturing samples increased with increasing test speed. The fracture surfaces were studied by scanning electron analysis, and sometimes a mixed mode fracture, part ductile and part brittle, could be seen. At high test speeds, a sharp brittle-ductile transition was observed, while at low test speeds the transition was more gradual, via a mixed mode region. This mixed mode region decreased in size with increasing test speed and was absent at the higher test speeds. The average crack speeds in the ductile region wer e directly related to the test speeds. The brittle-ductile transition temperature increased with the logarithmic of the test speed.


polycarbonate (PC) is a ductile polymer: however, when notched it often fails in a brittle manner [1-3], with a sharp brittle-ductile transition being observed as a function of the test temperature in the notched Izod or notched Charpy method [2, 4]. PC tested using the notched Izod method has a brittleductile transition near room temperature. This is well below its glass transition temperature. The rate of loading also has a great effect on the brittle-ductile transition [5-7], with a transition from ductile to brittle when the deformation rate is increased. These studies were carried out mainly at room temperature, and therefore the brittle-ductile transition temperatures of notched PC as a function of test speed have not yet been studied.

The brittle-ductile transition of polymers can be described by the Ludwing-Davidenkov-Ovowan criterion [8]. This criterion states that the fracture type changes from brittle to ductile if the yield stress drops below the fracture stress. With increasing temperature or decreasing test speed, the yield stress decreases more rapidly than the fracture stress. Consequently, the fracture type changes from brittle to ductile. With a ductile fracture, a considerable degree of plastic deformation takes place and most of this plastic deformation is dissipated as heat [9]. At high test speeds, the deformation process proceeds adiabatically, and a considerable temperature rise in the deformation zone can be expected. This adiabatic heating effect complicates the study of fracture as a function of test speed [10].

The mode of fracture in the ductile region is by ductile tearing, with tear lines visible at the fracture surface and the side walls being sucked in [11]. In the brittle region, the fracture is unstable, exhibiting a crazed breakdown pattern, and close to the notch, a mirror region can also be observed, followed by both mackerel and hackle regions [12]. Sometimes a mixed mode, partly ductile and partly brittle fracture behavior, is observed in polymers [6, 11, 12]. In this mixed mode deformation, the fracture starts to proceed in a ductile manner but abruptly becomes brittle. This type of behavior has been termed "ductile tearing instability" and is sometimes observed in PC, but as yet it is not clear under which conditions it is formed.

The influence of the test speed/strain rate on notched samples is usually studied by the Charpy method; however, an alternative method is the singleedge, notch tensile test (SENT). In the SENT set-up, the contact stiffness is much higher than in the Charpy method and the dynamic effects at high test speeds are not as extreme as in the Charpy method. The SENT method has for ductile fracturing polymers an additional advantage in that the samples are always fully broken.

This paper studies the brittle-ductile transition of notched polycarbonate specimens over a wide range of test speeds ([10.sup.-5] to 10 m/s) and temperatures (-80[degrees]C to +80[degrees]C), and the expected mixed mode region is defined. The test method used was the single-edged notch tensile (SENT).


Polycarbonate: Lexan 101-111 (granules) (a [M.sub.[omega]] of 29.500] kindly donated by General Electric Co., Bergen op Zoom, The Netherlands, was used. Prior to injection molding, the polymer was dried at 120[degrees]C for a minimum of four hours under a vacuum.

Injection molding: Rectangular bars (ISO 180/1, 74 X 10 X 4 mm) were injection molded on an Arburg 22 1-55-250 Allrounder injection molding machine. Starting from the hopper, the barrel temperatures were 290[degrees]C, 305[degrees]C, 310[degrees]C, and 315[degrees]C, with a nozzle temperature of 320[degrees]C and a screw speed of 100 rpm. The mold temperature was 850[degrees]C.

Notching: A single-edged, 45[degrees] V-shaped notch with a tip radius of 0.25 mm and a depth of 2.0 mm was milled into the bars.

Single-edged, notched tensile (SENT) testing: SENT tests were carried out using a Schenck VHS servo-hydraulic tensile machine [10]. All measurements were performed five-fold. With this apparatus, it is possible to achieve clamp speeds ranging from [10.sup.-5] to 12 m/s. The specimen length between the clamps was 34 mm, and therefore the macroscopic strain rate varied from 2.9 X [10.sup.-4] to 353 [S.sup.1].

On fracturing of the samples, crack initiation is assumed to occur at maximum load. Therefore, the point of maximum stress is chosen as the boundary between crack initiation and crack propagation. The following parameters are used to describe the fracture process:

- maximum stress:

Force maximum on the force-displacement curve, divided by the initial cross-sectional area behind the notch (32 [mm.sup.2]). Stress concentrations behind the notch are ignored.

- crack initiation energy:

Area under the force-displacement curve up to the force maximum.

- crack propagation energy:

Area under the force-displacement curve after maximum force.

- fracture energy:

Summation of the crack initiation and crack propagation energies.


The influence of test speed on the fracture behavior and in particular, on the brittle-ductile transition was studied using the SENT method. Using this instrumented test on notched samples, one is able to study both the fracture strength and the displacement. Figure 1 shows typical curves obtained a test speed of lm/s. Of the samples tested (at 1 m/s), the fracture was found to be brittle at -40[degrees]C and -15[degrees]C. At higher temperatures, the samples fractured in a ductile manner, with the fracture force being highest at 5[degrees]C. The fracture process can be divided into two stages, a deformation prior to crack initiation (crack initiation) and a deformation after crack initiation (crack propagation). During the initiation stage the stress builds up at the notch tip and some plastic deformation ahead of the notch can take place. At the maximum stress the crack propagation is assumed to start [10]. The sharp drop in the stress at low temperatures indicates that no additional energy is required for crack propagation. This type of fracture is referred to as a brittle fracture. At high temperatures, additional energy is required for crack propagation, and therefore this type of fracture is referred to as a ductile fracture.

The maximum stress is defined as the maximum force measured during the SENT test divided by the cross-sectional area behind the notch. The maximum stress is also the fracture stress. The maximum stress is given as a function of the test temperature at different test speeds (Fig. 2). At low temperatures, where all the samples fractured in a brittle manner, the maximum stress seems to decrease with increasing test speed. This suggests that at high test speeds, the stress concentration ahead of the notch is higher. At high temperatures, where all the samples fractured in a ductile manner, the maximum stress considerably increased with increasing test speed. In particular, the maximum stress values at 10 m/s were markedly higher than those at 1 m/s. This might be due to dynamic effects in the test [10, 13].

At every test speed there was, at some point in the curve, a strong increase in the total fracture energy (Fig. 3a). Up to the transition point, the fracture energy and the fracture displacement increased with increasing temperature. However, in the ductile region, the fracture energy stayed at the same level or even decreased with a further increase in test temperature. It is remarkable that the fracture energies in the high speed test are higher than in low speed tests.

As previously mentioned, the fracture process can be divided into two stages, where there is a deformation prior to crack initiation and a deformation following crack initiation, the crack initiation energy being the energy supplied to the sample prior to a crack being initiated. This crack initiation energy has been regarded as the parameter for the ductile-brittle transition [12, 14]. In these experiments the crack initiation energy varied with test speed but did have a maximum at the [] (Fig. 3b). In the brittle region the fracture energy increased, while in the ductile region, the fracture displacement seemed to decrease slightly with temperature. This implies that with an increasing temperature, the size of the ductile deformation zone ahead of the notch is not increased but is actually slightly decreased. This suggests that at higher temperatures the deformation is much more localized around the notch.

At high test speeds the crack propagation energy was found to increase in a stepwise manner (Fig. 3c), with the stepwise increase occurring at the same tem-perature as the maximum in the crack initiation curves (Fig. 3b). After this stepwise increase, the crack propagation energy was fairly constant.

At low test speeds the propagation energy with temperature increased over a wider temperature region, suggesting a gradual developing transition from brittle to ductile.

The higher total fracture energies in the ductile region (Fig. 3a) for the samples tested at high test speeds are due both to an increase in crack initiation energy (Fig. 3b) and in crack propagation energy (Fig. 3c). This must be due to the higher fracture stresses (Fig. 2) under these conditions. Possibly the adiabatic plastic deformation in the notch and ahead of a crack are playing a role in this.

From the fracture time and the crack length an average fracture speed can be calculated. In the mixed mode region of a partially ductile and partially brittle fracture this average speed is not meaningful. The fracture speed was determined as a function of the test temperature for a number of different test speeds (Fig. 4). Measuring a fast fracture after a slow loading proved to be difficult in the set-up used in these experiments. Therefore, the brittle fracture speeds were not measured in the low test speed experiments. In Fig. 3, the [] as obtained from the crack propagation energies is indicated by a vertical line.

At temperatures higher than the [], the fracture speed in the ductile region changed with the test speed. For the 1 m/s and the 10 m/s tests, the ductile fracture speeds were in the order of 5 m/s and 70 m/s, respectively. In the brittle region, the fracture speeds are expected to be very high and could only be determined of the high test speed tests due to a sampling rate in the test set-up. The fracture speeds were for the 1 m/s and 10 m/s tests in the region of 400-700 m/s. These values are considerably lower than the theoretical limiting value of the Raleigh wave speed ([v.sub.R]) of the material. This is probably due to crack branching that takes place at high crack speeds [15]. This crack branching limited the crack speed ([v.sub.c]) in PMMA to 0.4 [v.sub.R].

For the high test speeds, the change in fracture speed at the [] was found to be quite sharp. For low test speeds the fracture rate changed gradually at the []. The combination of a sharp brittle-ductile transition at high test speeds and a gradual transition at low test speeds has also been observed on PA-rubber blends [16] and PP-rubber blends [17].

The brittle-ductile transition in the SENT test was best determined from the crack propagation energy, with the jump in the fracture propagation energy being taken as the brittle-ductile transition (Fig. 3a). An additional way to determine the brittle-ductile transition is by studying fractured samples. A brittle fractured material has a mirror zone ahead of the notch followed by a heckled surface appearance, while a ductile fractured sample shows evidence of ductile tearing. If a mixed mode fracture takes place, then the brittle fracture is preceded by a ductile tearing zone, and this was observed at low temperatures and low test speeds (Fig. 5). At low test speeds the change from a fully brittle to fully ductile material occurred over a temperature interval, with the size of the ductile tearing zone increasing gradually with increasing temperature. At high test speeds, the transition from brittle to ductile was sharp. The points at which the material was fully brittle and fully ductile are presented in Fig . 6. This figure also shows the [] values as observed by SENT from the propagation energy curves. The [] as measured by SENT and the transitions observed in SEM photographs appear to correspond well. At low test speeds the transition to a fully brittle fracture was at much lower temperatures.

The temperature at the onset of a ductile fracture increased with the test speed, and a linear relationship of the [] with the log test speed was observed. At low test speeds the change from a fully brittle to a fully ductile material occurred over a temperature interval with a mixed mode deformation in this region. The size of the mixed mode region decreased with increasing test speed. The ductile tearing instability is thus only observed on samples fractured at low test speeds.


At the brittle-ductile transition, the fracture process changed from being a fast unstable fracture with craze formation to a slow stable fracture showing ductile tearing. At high test speeds this brittle-ductile transition is sharp, while at low test speeds, the transition is gradual and thereby goes through a mixed mode region having "ductile tearing instability." This phenomerion is not limited to PC. The size of this mixed mode region decreased with test speed and was absent at high test speeds. The [], for the onset of ductile fracture, increased linearly with the log test speed.

The fracture stress and the fracture energies in ductile fracturing samples increased with test speed. This effect might be partly due to adiabatic effects. The effect of adiabatic heating during plastic deformation on the [] is as yet unclear. However, it can be expected that if adiabatic heating ahead of the notch takes place, the yield stress is lowered very locally and strong deformation is limited to this warmed-up region. This yielding ahead of a notch might, however, blunt the notch, and as a consequence, the plastic deformation in the matrix would increase. These adiabatic effects need to be studied further.


This work was sponsored by FOM, Fundamental Materials Research Program of the Netherlands. The authors wish to thank L.C.E. Struik and Prof. E. van der Giessen for their support and stimulating discussions.

(*.)To whom correspondence should be addressed.

(**.)Currently at Philips. Drachten. the Netherlands.


(1.) R. A. W. Fraser and I. M. Ward, J. Mater. Sci., 12, 459 (1977).

(2.) F. C. Chang and L. H. Chu, J. Appl. Polym. Sci., 44, 1615 (1992].

(3.) G. Allen, D. C. W. Morley, and T. Williams, J. Mater. Sci., 8, 1499 (1973).

(4.) E. Platti and J. G. Williams, Polymer, 16, 915 (1975).

(5.) A. F. Yee, J. Mater. Sci., 12, 757 (1977).

(6.) M. E. J. Dekkers and S. Y. Hobbs, polym. Eng. Sci., 27, 1164(1987).

(7.) A. Kim, C. P. Bosnyak, and A. Chudnavky, J. Appl. Polym Sci., 51, 1841 (1994).

(8.) I. M. Ward, Mechanical Properties of Solid Polymers, Chapter 12, John Wiley & Sons (1971).

(9.) Y. K. Godovsky, Thermal Properties of Polymers, Chaper 8, Springer Verlag, New York (1992).

(10.) M.-P. Lee, A. Hiltner, and E. Baer, Polym. Eng. Sci., 32, 909 (1992).

(11.) W.-B. Liu and F.-C Chang, J. Appl. Polym. Sci., 56, 545 (1995).

(12.) K. Dijkstra, H. H. Wevers, and R J. Gaymans, Polymer, 35, 323 (1994).

(13.) Ph. Beguelin and H. H. Kauch, Impact and Dynamic Fracture of Polymers and Composites, p. 3, ESIS 19, J.G. Williams and A. Pavan, eds., Mechanical Engineering Publications, London (1995).

(14.) F.-C. Chang and H. C. Hsu, J. Appl. Polym. Sci., 52, 1891 (1994).

(15.) E. Sharon and J. Fineberg, Physical Review, 54, 7128 (1996).

(16.) R. J.Gaymans, K. Dijkstra, and M. H. ten Dam, in Toughend Plastics II, Chapter 20, p. 303, C. K. Riew and A. J. Kinloch, eds., ACS, Washington, Advances in Chemistry, no. 252 (1196).

(17.) A. van der Wal, A. J. J. Verheul, and R. J. Gaymans, Polymer, 40, 6045 (1999).
COPYRIGHT 2000 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:GAYMANS, R. J.; HAMBERG, M. J. J.; INBERG, J. P. F.
Publication:Polymer Engineering and Science
Geographic Code:1USA
Date:Jan 1, 2000
Previous Article:Effect of Incorporating Ethylene-Ethylacrylate Copolymer on the Positive Temperature Coefficient Characteristics of Carbon Black Filled HDPE Systems.
Next Article:Carboxyl Terminated Polyamide 12 Chain Extension by Reactive Extrusion Using a Dioxazoline Coupling Agent. Part I: Extrusion Parameters Analysis.

Related Articles
Measurement of the J integral in polycarbonate using the method of caustics.
Ballistic performance of polycarbonate/polyester and polycarbonate/styrene-acrylonitrile microlayer sheets.
Determination of fracture toughness in rubber modified glassy polymers under impact conditions.
Scratch hardness and deformation maps for polycarbonate and polyethylene.
Polymer Toughness and Impact Resistance.
Effects of Gauge Length and Strain Rate on Fracture Toughness of Polyethylene Terephthalate Glycol (PETG) Film Using The Essential Work of Fracture...
Influence of Low Molecular Weight ABS Species on Properties of PC/ABS Systems.
A New Method to Quantify Crazing in Various Environments.
Effect of Entanglement on Brittle-Ductile Transition in Polyethylene.
Structure-property relationships of blends of polycarbonate.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters