Unlike crystalline metals, which can accommodate plastic strain through slip, twinning, etc., bulk metallic glasses undergo permanent deformation through formation of shear bands. In most BMGs, plastic deformation is a function of sample geometry and experimental testing conditions [1]. For instance, Inoue et al. [2] and Katsuya et al. [3] found that as sample size increases, bending ductility decreases. Conner et al. [1, 4] showed that as sample thicknesses approach the “bulk” level (~ 1 mm), there is a significant reduction in the number of shear bands generated, and thus, less plastic strain prior to failure.
To avoid catastrophic failure in unconstrained loading (as might be experienced in engineering applications) future BMGs or their composites must be able to exhibit plastic strain in tension and bending. Thus far, the only method for creating an alloy that is ductile in tension testing is through the use of a composite structure, where a softer dendritic phase inhibits shear band movement [5, 6] (see Chapter 4). However, recently a new class of alloys has been developed, based on the simple Cu-Zr binary system, that shows high levels of plastic strain in uniaxial compression tests [7–13].
Three-point-bending tests are rarely used in literature involving BMGs and yet they are quite useful for predicting the mechanical behavior of glasses in semi- unconstrained loading. Since tension tests always result in catastrophic failure in BMGs, and compression tests only give plastic strain (not ductility), bending tests provide a unique compromise between the two.
3.3
Exper imental Method
The alloys used in this work were first prepared as an ingot by arc melting ultrasonically cleansed 99.999 at.% Cu, 99.9 at.% Zr, and 99.9 at.% Al on a water cooled copper plate in a high-purity argon environment. Each ingot was melted three times to promote homogeneity. The ingots were then melted in a quartz tube through an induction coil and injected with pressurized argon (0.2–0.4 MPa) into a copper mold. The cavity has thickness 1 mm, width 3 mm, and length 40 mm. Bending specimens were cut from the strip to a length of 25 mm and then polished with 3 micron diamond suspension. Three-point bending was completed on an Instron 5500R load cell with the sample resting on two 6.35 mm diameter pins separated by 18.89 mm and bent by a third pin of the same diameter. TEM work was done on a Phillips EM420 at 120 keV using samples that were prepared electrochemically with 25% nitric acid in methanol.
3.4
Results of Bending Tests
Figure 3.1 shows bending load versus bend displacement curves for five 1-mm- thick Cu-Zr based alloys bent in the three-point-bend configuration, shown in the inset. Each alloy shows a non-linear effect at small displacements that is caused by machine compliance. Only two alloys, Cu47.5Zr47.5Al5 and Cu16.4Zr57.4Ni8.2Ta8Al10, show a
significant amount of plastic strain prior to failure at this thickness, indicating that they have enhanced fracture resistance.
Optical micrographs of shear bands on the tensile side of four bending specimens are shown in Figure 3.2. Each micrograph shows the shear band formation
near the point of fracture. Horizontal lines in the micrographs are damage from polishing and not from deformation.
Figure 3.3 is an example of the evolution of mixed mode cracks across the tensile side of the bending surface in Cu47.5Zr47.5Al5, and is used to estimate the bending
displacement at the onset of plastic strain. The sample was polished and then loaded in steps to increasing displacements. Following each step, the sample was removed from the testing equipment and viewed via optical microscopy for the presence of visible deformation. Figures 3.3(a–b) are micrographs of the sample bent to 1.25 mm and 1.50 mm, respectively. Just prior to 1.50 mm of displacement, mixed mode cracks begin to form and propagate from the edge of the sample across the width, indicating the onset of plastic strain. This coincides with the onset of the non-linear curvature in Figure 3.1 for Cu47.5Zr47.5Al5.
Figure 3.3(c) is a higher magnification image of the arrow in Figure 3.3(b). The micrograph shows a mixed-mode crack at the surface being blunted by several small shear bands at its tip. From Figure 3.3(b) this plastic region in front of the mixed-mode crack extends approximately 450–600 μm into the sample. This plastic zone size can be used to approximate the fracture toughness of the glass (see Section 3.7). Figure 3.4 is an optical micrograph of the side view of a 1 mm Cu47.5Zr47.5Al5 bending sample. The
shear band spacing and offsets are estimated to be approximately 180 μm and 15.4 μm, respectively.
Prior to bending, the microstructure of Cu47.5Zr47.5Al5 was evaluated through
transmission electron microscopy (TEM). TEM foils were prepared from a 1-mm-thick bending specimen and then viewed less than five minutes after being chemically
thinned. The SAED pattern (Figure 3.5(c)) from a nominal part of the sample demonstrates that Cu47.5Zr47.5Al5 has a fully amorphous structure. Two broad diffuse
rings, characteristic of an amorphous material, are apparent in the diffraction pattern and no crystallization is observed in bright field or dark field imaging. To determine the effect of oxidation on the TEM images, the foils were stored in air for three days and viewed again for comparison. Figures 3.5(a–b) demonstrate that the oxidation of the sample now causes it to appear crystalline. The bright field image displays surface discoloration, the dark field image has small diffracting crystallites, and the SAED pattern has a thin, crystalline halo.
Figure 3.1 – Bending load versus bend displacement plot for five alloys bent in three- point-bending configuration shown in the inset. Despite all having an amorphous microstructure, some BMGs exhibit higher resistance to fracture in bending tests, as evidenced by their bending plasticity.
Figure 3.2 – Tensile surfaces of bending samples in (a) Cu45Zr45Ti10, (b)
Cu16.4Zr57.4Ni8.2Ta8Al10, (c) Cu46Zr45Al7Y2, (d) Cu47.5Zr47.5Al5.
Figure 3.3 – (a) Cu47.5Zr47.5Al5 loaded to 1.25 mm of bend displacement and then
unloaded; (b) loaded to 1.50 mm showing onset of plastic strain. (c) enlargement of arrow showing that the plastic region in front of a crack is comprised of many small shear bands. The size of the plastic zone can be estimated from (b) to be several hundred microns.
Figure 3.4 – Side view of a bending sample of Cu47.5Zr47.5Al5 showing shear band
offsets and shear band spacing. The sample was unpolished on the sides so horizontal lines are flow lines from casting, not from deformation. When the shear band offset become too large (at some specific shear band length) cracks nucleate and the beam fractures.
Figure 3.5 – (a) BF of Cu47.5Zr47.5Al5 after three days exposed to air, showing
discoloration and apparent crystallization. (b) DF showing small crystallites and a crystalline halo in the inset SAED pattern. (c) SAED pattern for Cu47.5Zr47.5Al5 showing
Figure 3.6 – Geometry of a sample in pure bending.
Figure 3.7 – Bending load versus bending displacement curve for Cu47.5Zr47.5Al5,