Ductile crystalline–amorphous nanolaminates

Wang et al. 10.1073/pnas.0702344104.

Supporting Information

Files in this Data Supplement:

SI Figure 5
SI Text
SI Figure 6
SI Figure 7
SI Figure 8
SI Movie 1
SI Movie 2
SI Movie 3

SI Figure 5

Fig. 5. Cutting die. Two stainless steel plates with the sample shape were used for machining of tensile samples reported in this work.

SI Figure 6

Fig. 6. MD simulation system setup for 5/10 nanolaminates. The yellow and gray spheres represent copper and zirconium atoms, respectively. Periodic boundary condition was applied in all three directions. The tensile direction is along

SI Figure 7

Fig. 7. (A) MD simulations at T = 300 K shows dense forest of sessile dislocations formed soon after the onset of the plastic flow in 5/10 system. (B) However, when an additional plastic strain of 10% was sustained, the sessile dislocation density dropped to zero, and only one or two mobile threading dislocations were left. Atoms are color-coded by their coordination numbers [red, 11; blue, 13; green, 10; tan, 14; etc.; perfectly coordinated atoms (12) are not rendered]. SI Movie 2 shows the same destabilization process, without filtering out the thermal noises.

SI Figure 8

Fig. 8. The tensile stress-strain curves of polycrystalline 5/10 (A) and 5/35 (B) Cu/Zr systems from MD simulations at the strain rate of ≈1 ´ 10⁸ s^-1. Because the simulation starts from a perfect dislocation-free structure, a large stress is required to nucleate dislocations from grain boundaries and amorphous-crystalline interfaces. After 5-7% strain, a steady-state flow stress is established. The lower peak stress (1.7 GPa versus 1.8 GPa) but higher steady-state flow stress (1.4 GPa versus 1.2 GPa) of the 5/10 system compared with the 5/35 system is characteristic of the MD simulation behavior. The artificially high peak stress of the 5/35 simulation is due to the pristine initial sample condition and the strong sensitivity of nucleation stress to strain rate.

SI Movie 1

Movie 1. Inelastic sheer strain evolution in 5/10 nanolaminate.

SI Movie 2

Movie 2. Dislocation structure evolution (longer term) in 5/10 nanolaminate

SI Movie 3

Movie 3. Dislocation structure evolution (longer term) in 5/35 nanolaminate.

SI Text

SI Fig. 6 shows the 5/10 system setup (gray sphere, Zr; yellow sphere, Cu) for simulations. Uniaxial tension is applied in <112> direction, with the strain rate of ≈1 ´ 10⁸ s^-1. Periodic boundary conditions are applied in all three dimensions. Dislocation nucleation, absorption, activation of STZ, and gradual blunting of stress concentrations near ACIs can be seen in SI Movie 1, where the inelastic or transformation strain was calculated and rendered atom by atom (the same scale bar as Fig. 3D). Those atoms with the inelastic strain <6% are not shown.

The initial creation and subsequent destabilization of the sessile dislocation forests with the progress of supercell strain (0-15%) are shown in SI Fig. 7 and SI Movie 2. Coordination number coloring is used in both the figure and the movie. In SI Movie 2, thermal noises (T = 300 K) are not filtered out. The mobile and sessile dislocations are seen to be drawn into the amorphous layer. No dislocation entangles were left at the end of the simulation, but only one or two mobile threading dislocations. SI Movie 3 shows dislocation structure evolution in the 5/35 polycrystalline nanolaminate with increasing strains from 0% to 15%. It is evident that the presence of the amorphous layer strongly attracts and destabilizes dislocation structures nearby. No shear bands were formed in the amorphous layers during MD simulations despite the high strain rates. Deformation twinning was often observed in the nanocrystalline layers. SI Fig. 8 show tensile stress-strain curves of 5/10 and 5/35 systems from MD simulations.