Page 350 - T. Anderson-Fracture Mechanics - Fundamentals and Applns.-CRC (2005)

1656_C007.fm Page 330 Monday, May 23, 2005 5:54 PM

330 Fracture Mechanics: Fundamentals and Applications

Because high-rate fracture tests typically last only a few milliseconds, conventional data acqui-
sition tools are inadequate. A storage oscilloscope has traditionally been required to capture data
in a high-rate test; when a computer data acquisition system was used, the data were downloaded
from the oscilloscope after the test. The newest generation of data acquisition cards for PCs removes
the need for this two-step process. These cards are capable of collecting data at high rates, and
enable the computer to simulate the functions of an oscilloscope.
Inertia effects can severely complicate the measurement of the relevant fracture parameters.
The stress-intensity factor and J integral cannot be inferred from global loads and displacements
when there is a significant kinetic energy component. Optical methods such as photoelasticity and
caustics are necessary to measure J and K in such cases.
The transition time concept [25,26], which was introduced in Chapter 4, removes much of the
complexity associated with J and K determination in high-rate tests. Recall that the transition time
t is defined as the time at which the kinetic energy and deformation energy are approximately
τ
equal. At times much less than t , inertia effects dominate, while inertia is negligible at times
τ
significantly greater than t . The latter case corresponds to essentially quasistatic conditions, where
τ
conventional equations for J and K apply. According to Figure 4.4, the quasistatic equation for J,
based on the global load-displacement curve, is accurate at times greater that 2t . Thus, if the
τ
critical fracture event occurs after 2t , the toughness can be inferred from the conventional quasi-
τ
static relationships. For drop tower tests on ductile materials, the transition time requirement is
relatively easy to meet [27,28]. For brittle materials (which fail sooner) or higher loading rates,
the transition time can be shortened through specimen design.
7.6.2 K MEASUREMENTS
Ia
In order to measure arrest toughness in a laboratory specimen, one must create conditions under
which a crack initiates, propagates in an unstable manner, and then arrests. Unstable propagation
followed by arrest can be achieved either through a rising R curve or a falling driving force curve.
In the former case, a temperature gradient across a steel specimen produces the desired result;
fracture can be initiated on the cold side of the specimen, where toughness is low, and propagate
into warmer material where arrest is likely. A falling driving force can be obtained by loading the
specimen in displacement control, as Example 2.3 illustrates.
The Robertson crack-arrest test [29] was one of the earliest applications of the temperature-
gradient approach. This test is only qualitative, however, since the arrest temperature, rather than
K , is determined from this test. The temperature at which a crack arrests in the Robertson specimen
Ia
is only indicative of the relative arrest toughness of the material; designing above this temperature
does not guarantee crack arrest under all loading conditions. The drop weight test; developed by
Pellini (see Section 7.9) is another qualitative arrest test that yields a critical temperature. In this
case, however, arrest is accomplished through a falling driving force.
While most crack-arrest tests are performed on small laboratory specimens, a limited number
of experiments have been performed on larger configurations in order to validate the small-scale
data. An extreme example of large-scale testing is the wide plate crack-arrest experiments conducted
4
at the National Institute of Standards and Technology (NIST) in Gaithersburg, Maryland [30].
Figure 7.30 shows a photograph of the NIST testing machine and one of the crack-arrest specimens.
This specimen, which is a single-edge notched tensile panel, is 10 m long by 1 m wide. A
temperature gradient is applied across the width, such that the initial crack is at the cold end. The
specimen is then loaded until unstable cleavage occurs. These specimens are heavily instrumented,
so that a variety of information can be inferred from each test. The crack-arrest toughness values
measured from these tests are in broad agreement with small-scale specimen data.

4 NIST was formerly known as the National Bureau of Standards (NBS), which explains the initials on either end of the
specimen in Figure 7.30.