5.1 Tensile-impact energy is the energy required to break a standard tension-impact specimen in tension by a single swing of a standard calibrated pendulum under a set of standard conditions (see Note 2). To compensate for the minor differences in cross-sectional area of the specimens, the energy to break is normalized to units of kilojoules per square metre (or foot-pounds-force per square inch) of minimum cross-sectional area. An alternative approach to normalizing the impact energy that compensates for these minor differences and still retains the test unit as joules (foot-pounds) is shown in Section 10. For a perfectly elastic material, the impact energy is usually reported per unit volume of material undergoing deformation. However, since much of the energy to break the plastic materials for which this test method is written is dissipated in drawing of only a portion of the test region, such normalization on a volume basis is not feasible. In order to observe the effect of elongation or rate of extension, or both, upon the result, the test method permits two specimen geometries. Results obtained with different capacity machines generally are not comparable.

5.1.1 With the Type S (short) specimen the extension is comparatively low, while with the Type L (long) specimen the extension is comparatively high. In general, the Type S specimen (with its greater occurrence of brittle fracture) gives greater reproducibility, but less differentiation among materials.

Note 2-Friction losses are largely eliminated by careful design and proper operation of the testing machine.

5.2 Scatter of data is sometimes attributed to different failure mechanisms within a group of specimens. Some materials exhibit a transition between different failure mechanisms. If so, the elongation will be critically dependent on the rate of extension encountered in the test. The impact energy values for a group of such specimens will have an abnormally large dispersion.

5.2.1 Some materials retract at failure with insignificant permanent set. With such materials, determining the type of failure, ductile or brittle, by examining the broken pieces is difficult, if not impossible. It is helpful to sort a set of specimens into two groups by observing the broken pieces to ascertain whether or not there was necking during the test. Qualitatively, the strain rates encountered here are intermediate between the high rate of the Izod test of Test Methods D256 and the low rate of usual tension testing in accordance with Test Method D638.

5.3 The energy for fracture is a function of the force times the distance through which the force operates. Therefore, given the same specimen geometry, it is possible that one material will produce tensile-impact energies for fracture due to a large force associated with a small elongation, and another material will produce the same energy for fracture result due to a small force associated with a large elongation. It shall not be assumed that this test method will correlate with other tests or end uses unless such a correlation has been established by experiment.

5.4 Comparisons among specimens from different sources are to be made with confidence only to the extent that specimen preparation, for example, molding history, has been precisely duplicated. Comparisons between molded and machined specimens must not be made without first establishing quantitatively the differences inherent between the two methods of preparation.

5.5 Only results from specimens of nominally equal thickness and tab width shall be compared unless it has been shown that the tensile-impact energy normalized to kilojoules per square metre (or foot-pounds-force per square inch) of cross-sectional area is independent of the thickness over the range of thicknesses under consideration.

5.6 The bounce of the crosshead supplies part of the energy to fracture test specimen (see Appendix X1).

5.7 For many materials, there are specifications that require the use of this test method, but with some procedural modifications that take precedence when adhering to the specification. Therefore, it is advisable to refer to that material specification before using this test method. Table 1 of Classification System D4000 lists the ASTM materials standards that currently exist.