Dependence of strain rate sensitivity upon deformed microstructures in nanocrystalline Cu

Abstract

The creep behavior of nanocrystalline copper was experimentally characterized with nanoindentation using two sequential regimes, i.e., loading and holding. Significantly enhanced strain rate sensitivity was found within an unusually narrow range of creep rate in the holding regime, which is attributed to the deformed microstructure generated during the loading regime. By quantitatively analyzing the creep rate and rate sensitivity exponent of NC Cu in the holding regime, both the grain boundary sliding (GBS) and dislocation activities are found to be responsible for the observed abnormal behavior, with the contribution of GBS decreasing with increasing grain size and increasing with decreasing loading strain rate. These findings provide a potential way of adjusting the mechanical properties of nanocrystalline metals by pre-straining.

Introduction

The nanoscale grain size in nanocrystalline (NC) metals results in plastic deformation mechanisms that are substantially different from those in coarse-grained (CG) metals. Additionally, the confined nanoscale grain structure plays an important role in the mechanical properties of NC metals [1], [2]. It has been demonstrated that molecular-dynamics (MD) simulations, which can reflect the effects of both grain size and grain structure, are effective for in situ analysis of the deformation mechanisms of NC metals at atomic level [3], [4], [5], [6]. However, MD simulation can only capture the very beginning of the deformation process owing to its sub-nanosecond time scale [1]. Its generalization to real deformation processes should be considered with discretion [2], and the findings derived from MD simulations need to be validated with experiment measurements. In practice, it is not an easy task to control the grain structures in NC metals. The confined grain structure and the large volume fraction of the grain boundary (GB) make the interplay between GB and dislocation more complicated, and hence highlighting the intricate interplay of the microstructures in the lab environment is difficult. As a result, in general, the grain size has been proposed as the sole key parameter to characterize the mechanical properties of NC metals, although it is known that this is a less effective way than originally expected.

Such simplifications would, in some cases, give rise to uncertainties and controversies. For a CG metal, the grain size dependence of its hardness is consistent with the classical Hall–Petch (H–P) law. When the grain size is reduced to below a critical value at nanoscale, both experiments and MD simulations have revealed that the H–P relation breaks down, with GB activities dominating the plastic deformation [1], [7], [8]. To understand the transition of deformation mechanism associated with varying grain size, considerable efforts have been made to explore quantitatively the critical grain size. Nonetheless, controversies still persist. With NC Cu as an example, MD simulations predict that GB activities will dominate the plastic deformation when the grain size is less than ∼10–15 nm [9]. However, existing experimental studies only show a plausible trend of hardness when the grain size approaches the predicted critical value [10]. Moreover, the H–P law was found to persist even when the grain size was reduced to 10 nm [11]. Besides this discrepancy between experiments and predictions, the MD simulation itself also results in inconsistent findings. A recent MD simulation predicted that dislocations could contribute ∼50% of the plastic strain in Cu with a grain size of 5 nm [12], which is much smaller than the well-documented predictions of the critical grain size of Cu [9], [13].

Other than MD simulations, transmission electron microscopy (TEM) is another effective technique for evaluating the evolution of microstructures during deformation [14], [15], [16], [17]. This, however, also has some uncertainties. For in situ TEM study of ultra-thin TEM foils, the observation may be hindered by enhanced diffusion from nearby free surfaces, leading to possible wrong interpretations of the observed diffraction contrast [7]. Ex situ TEM observations, in contrast, may be flawed by stress relaxation during specimen preparation. Furthermore, as the ex situ TEM can only reveal the final state of the deformation, understanding the entire deformation process is difficult. Accordingly, the link between the confined nanoscale grain structure and the mechanical properties has remained open in NC metals.

Given that GB acts as both a source and a sink for dislocations that extend throughout the entire grain in NC metals [1], [2], Carlton and Ferreira [10] recently proposed a model for the inverse H–P effect, with a new parameter, the probability of a dislocation absorbed by the GB, introduced. This model potentially demonstrates a way to describe the local grain structure environment and stressing state during deformation. Based on Carlton’s work, the present authors propose that a test consisting of two sequential deformation processes might be able to uncover the link between the internal grain structures of a NC metal and its mechanical properties. While the first process sets up an environment for various deformed microstructures to evolve, the following one aims to demonstrate the features of the deformed microstructures generated during the preceding process. To this end, nanoindentation creep is considered a suitable testing method. In the present study, the nanoindentation creep test has two separate regimes: the loading regime in which the load is continuously increased, and the holding regime in which the load remains constant. For the loading regime, Carlton’s model can be adopted, in which the ability to absorb dislocations is changed by varying the loading strain rate (LSR). For the holding regime, as the extended dislocations will be trapped inside a nanoscale grain under very high stresses [18], the giga-pascal level stresses induced by the stabilized stresses underneath the nanoscale indenter prevent the emitted dislocations from running back to where they originally nucleate. Furthermore, as both the emission and absorption processes can alter the grain structures by influencing the local GB and dislocation networks in NC metals [3], [4], [5], [6], the deformed microstructures generated in the loading regime can be demonstrated by examining the creep behavior in the holding regime.

In the present experimental study, the loading regime is employed as a pre-straining process to generate different deformed microstructures at different LSR, and the holding regime is designed to validate the proposed model. The strain rate dependence of the hardness and creep behavior of NC Cu is quantified to explore the link between its deformed microstructure and mechanical response.