The mechanical testing technique for in situ nanoindentation in a transmission electron microscope is described and is shown to provide real-time observations of the mechanisms of plastic deformation that occur during nanoindentation. Here, the importance of this technique was demonstrated on an aluminum thin film deposited on a single-crystalline silicon substrate. Significant results include direct observation of dislocation nucleation, characterization of the dislocation distribution created by indentation, and the observation of indentation-induced grain boundary motion. The observations achieved by this technique provide unique insight into mechanical behavior studied with conventional instrumented nanoindentation techniques and also provide microstructural-level understanding of the mechanics of ultrasmall volumes.

The distinctive finding of unstable displacement bursts of indent depth observed in nanoindentation is thought to be linked to the dislocation emission phenomenon. However, because of a lack of understanding of the surface physics of the contact area, the explanation of the instability mechanism remains vague. In this paper, an effect of the surface roughness on the displacement bursts was experimentally investigated using single-crystalline aluminum. The surface properties of samples with several degrees of roughness of less than 10 nm were measured using atomic force microscopy. The discussion focuses on the relations between critical values of the load at the displacement bursts and the burst width.

The influence of grain boundaries on material deformation in Ni3Al was investigated by relating the material pile-up at grain boundaries and the propagation of slip across grain boundaries to the misorientation between the corresponding grains. Indentation tests were carried out using micro- and nanoindentation at distances shorter than the radius of indent size from a grain boundary on Ni3Al. The indents were observed using scanning electron microscopy and non-contact-mode atomic force microscopy. Repeated experimentation did not reveal a rising trend of hardness near grain boundaries, indicating that hardness is not a sensitive parameter to measure grain boundary strengthening effects. However, it was observed that the slip transfer behavior across a grain boundary has a strong dependence on a local misorientation factor m′ relating the misorientation of slip planes and slip directions on either side of the grain boundary. This result agrees with the fundamental assumption in the physical explanation of the Hall–Petch effect.

The surface layer of A319 alloy was refined by selective remelting using laser energy. Instrumented indentation technique was used to measure hardness (H) and elastic modulus (E) of the laser-melted layer. Berkovich tip was used to indent the material for 100-nm, 200-nm, 500-nm, and 1000-nm depths. The H and the E were found to be 1.22 GPa and 78.2 GPa, respectively, for 1000-nm indentation depths. The variances associated with H and E were minimal, whereas, the same for substrate possessed significant scattering. Also, H and E increased with decreasing depth of indentation. Closer examination suggested that when the heterogeneity in the material was in the scale of indentation depth, significant scattering took place and the hard phase Si influenced the average hardness. However, the influence of indentation depth on elastic modulus was not statistically significant.

Nanoindentation test at scale of hundreds of nanometers has shown that measured hardness increases strongly with decreasing indent depth, which is frequently referred to as the size effect. Usually, the size effect is displayed in the hardness-depth curves. In this study, the size effect is characterized in both the load–displacement curves and the hardness–depth curves. The experimental measurements were performed for single-crystal copper specimen and for surface-nanocrystallized Al-alloy specimen. Moreover, the size effect was characterized using the dislocation density theory. To investigate effects of some environmental factors, such as the effect of surface roughness and the effect of indenter tip curvature, the specimen surface profile and the indentation imprint profile for single-crystal copper specimen were scanned and measured using the atomic force microscopy technique. Furthermore, the size effect was characterized and analyzed when the effect of the specimen surface roughness was considered.

Thin-film mechanical properties can be measured using nanoindentation combined with detailed finite element modeling. This technique was used for a study of very fine grained Ni films, formed using pulsed-laser deposition on fused silica, sapphire, and Ni substrates. The grain sizes in the films were characterized by electron microscopy, and the mechanical properties were determined by ultra-low load indentation, analyzed using finite element modeling to separate the mechanical properties of the thin layers from those of the substrates. Some Ni films were deposited at high temperature or annealed after deposition to enlarge the grain sizes. The observed hardnesses and grain sizes in these thin Ni films are consistent with the empirical Hall–Petch relationship for grain sizes ranging from a few micrometers to as small as 10 nm, suggesting that deformation occurs preferentially by dislocation movement even in such nanometer-size grains.

The elastoplastic surface deformation of a polycrystalline graphite was studied by examining the indenter’s geometry dependence of load P versus penetration depth h relation (P–h relation) in instrumented pyramidal/spherical indentation tests. The tetrahedral pyramid indenters included inclined face angles β of 10.0°, 22.0° (Vickers pyramid), and 40.0°. The tip radius of spherical indenters used were 32 μm, 200 μm, 794 μm, 1.59 mm, and 6.35 mm. The true hardness H as a measure for plasticity was singled out of the elastoplastic loading parameter k1 in the quadratic expression of P = k1h2 and then quantitatively related to the yield stress Y that was determined from the mean contact pressure for spherical indentation at the onset of plastic yielding. The size effect of Y, decreasing with the increase in the tip radius of spherical indenter, is discussed using the model of geometrically necessary dislocations in terms of the material length scales for a plastic field with strain gradient.

The deformation of polymer films under repetitive contact at high strain rates was investigated using nanoscale impact testing. Four systems were studied: (i) rubber-modified acrylonitrile-butadiene-styrene (ABS) (0–25 wt% rubber), (ii) uniaxially and biaxially drawn poly(ethylene terephthalate) film; (iii) poly(ethylene oxide)–clay nanocomposites, and (iv) nylon 6–organoclay nanocomposites. The initial results suggest that the technique has much potential in evaluating the fatigue behavior of thinner polymer films and coatings that are unsuitable for conventional methods designed for bulk samples. The extent of impact-induced deformation may be used as a measure of ductility because ductile failures are associated with significant plastic deformation before failure whereas brittle failures usually involve little plastic deformation. The nano-impact technique provides valuable highly localized information about deformation under high strain rate, which is complementary to low strain rate tests such as nanoindentation and nano-scratch. The technique has been shown to be sensitive to nano-/microstructural variations in ABS–rubber film when Berkovich indenters and low impact forces were used. The impact behavior of the nanocomposites is only significantly worse than that of the virgin polymers at the highest clay loading studied (15 wt%). This could be a factor when assessing the suitability of novel nanocomposite materials for applications where toughness is important. On ABS film, there is only an approximate correlation between the plastic work function determined from nanoindentation and the rubber loading in the film while the correlation between the rubber loading and nano-impact data is clear, suggesting that the dynamic test is a more useful predictor of thin polymer film toughness than the slow-loading quasi-static indentation test.

Elastic modulus of bone was investigated by nanoindentation using common methods of sample preparation, data collection, and analysis, and compared to dynamic mechanical analysis (DMA: three-point bending) for the same samples. Nanoindentation (Berkovich, 5 μm and 21 μm radii spherical indenters) and DMA were performed on eight wet and dehydrated (100% ethanol), machined equine cortical bone beams. Samples were embedded in polymethylmethacrylate (PMMA) and mechanical tests repeated. Indentation direction was transverse to the bone long axis while DMA tested longitudinally, giving approximately 12% greater modulus in DMA. For wet samples, nanoindentation with spherical indenters revealed a low modulus surface layer. Estimates of the volume of material contributing to elastic modulus measurement showed that the surface layer influences the measured modulus at low loads. Consistent results were obtained for embedded tissue regardless of indenter geometry, provided appropriate methods and analysis were used. Modulus increased for nanoindentation (21 μm radius indenter) from 11.7 GPa ± 1.7 to 15.0 GPa ± 2.2 to 19.4 GPa ± 2.1, for wet, dehydrated in ethanol, and embedded conditions, respectively. The large increases in elastic modulus caused by replacing water with ethanol and ethanol with PMMA demonstrate that the role of water in fine pore space and its interaction with collagen strongly influence the mechanical behavior of the tissue.

Depth-sensing indentation was applied to three-dimensionally ordered silica foams of two different pore diameters—500 nm and 850 nm—formed by colloidal crystal templating. The contact responses of indentations with Berkovich and hemispherical indentation tips are presented over a load range of 1 mN to 100 mN. Scanning electron microscopy images of residual indentation impressions showed homogeneous deformation for small loads in which the peak displacement was shallow relative to the film–substrate interface. The characteristics of the load–displacement responses changed from periodic discontinuities, associated with cell wall fracture and pore collapse, to smooth and increased stiffness, as a result of densification due to the accumulation of material under the indentation tip and proximity (and contact) of the substrate. Load–displacement responses were translated into pressure–volume space, in which the average pressure during indentation is a measure of the crushing pressure of the cell walls.

Cartilage repair is an ongoing challenge in orthopaedic research. Both biological and mechanical properties are needed for full functional characterization of a tissue replacement. However, assessment of cartilage repair tissue has relied primarily on histological markers because mechanical properties are difficult to measure in small tissue regions. Nanoindentation is a technique that is specialized for measuring local mechanical properties in small specimens. In this study, nanoindentation was used to measure the mechanical properties of cartilage repair tissue following microfracture treatment of chondral defects in three rabbit knees. The mechanical properties were then correlated with histology findings from the same tissue samples. Nanoindentation provided complementary information to histology. In particular, the nanoindentation results showed that the mechanical properties of the repair tissue were inferior to the control tissue properties, even though some tissue regions were histologically similar to the control tissue.

The ability of using nanoindentation raw data for characterizing the strain rate sensitivity of the flow stress of elastic–plastic materials has been assessed by finite element calculations. Correction factors for deducing the true strain rate sensitivity from the hardness sensitivity to the hardness penetration rate have been obtained for situations where the elastic contribution to the penetration is not negligible. The results of the analysis are applied to the experimental measurement by nanoindentation of the strain rate sensitivity of an ultrafine grain Sn–Bi eutectic processed by equal channel angular pressing. The flow stress of this material displays bilinear power-law strain-rate dependence. It is also shown that at very low indentation depths, discontinuities of the hardness–penetration curve of the fine-grain Sn–Bi can be correlated with local grain or interphase boundary slip events near the indentation.

An iron aluminide (Fe–40 at.% Al) was thermal oxidation (TO)-treated at 1000 °C for 30 to 150 h. The mechanical properties (e.g., hardness and elastic modulus), the scratch resistance, the deformation and damage behaviors of the TO-treated surface layers were evaluated using nanoindentation and nanoscratch techniques. It has been found that the TO treatment produced an oxide layer containing α–Al2O3 on the iron aluminide surface, which increased the surface hardness (H) and elastic modulus (Er) but reduced the ratio of Er/H. This has significantly improved the scratch resistance of the iron aluminide. Longer treatment resulted in a thicker oxide layer and led to further reduced scratch rate and an increased critical load to failure. The scratch resistance has been correlated with the scratch load, the hardness, and the Er/H ratio of the surface oxide layer.

The determination of the mechanical properties of ultra-thin coatings has become more and more important because of the increasing number of applications using such films. However, an accurate mechanical testing of coatings with a thickness down to some nanometers is still a challenge, despite the improvements of existing measurement techniques. Nanoindentation is an often used mechanical nanoprobe. Using the conventional test method with a sharp Berkovich indenter, the problem of the influence of the substrate on the results arises with decreasing film thickness. Therefore, it is nearly impossible to measure the modulus of films with a thickness less than 100–200 nm. The problem can be overcome by using spherical indenters in combination with an analytical solution for the Hertzian contact of coated systems. It allows a separation of film and substrate properties from the load–displacement curve of the compound. Indentation measurements were done at a 44 nm TiN film and at diamondlike carbon coatings in the thickness range between 4.3 nm and 125 nm on Si substrates. Several corrections were applied to obtain wholly elastic force–displacement curves with high accuracy. It is shown in more detail how zero point and thermal drift corrections are used to obtain statistical depth errors below 0.2 nm. Laser-acoustic measurements based on ultrasonic surface waves were chosen as a second method, which also measures the Young’s modulus in this thickness range. Although the indentation technique is a local probe and the laser-acoustic technique gives an integrated value for a surface range of some millimeters, the results agree well for the investigated samples. In contrast, it was impossible to get the correct Young’s modulus results by conventional indentation measurements with Berkovich indenter, even for ultra-low loads.

A comprehensive study was undertaken to identify the extent to which the mechanical properties of thin metal films on substrates could be determined quantitatively from instrumented sharp indentation. The mechanical behavior of thin Cu films on substrates was investigated using three different methods: nanoindentation, cantilever microbeam deflection, and microtensile testing. Finite element calculations of the nanoindentation and microbeam deflection experiments were conducted to extract yield strength and hardening modulus. Systematic experiments were performed to investigate the consistency of the different experimental techniques. The mechanical behavior of the Cu films was observed to depend on the film thickness. However, the results from finite element modeling of nanoindentation and microbeam deflection are quite different. In both cases, unique solutions for yield strength and hardening modulus were found. This is particularly noteworthy for the nanoindentation experiments; it is argued that the substrate destroys the self-similarity that is present during indentation of bulk material using a Berkovich tip. Microbeam deflection experiments seem to be more sensitive to the elastic–plastic transition, whereas the nanoindentation results describe the mechanical behavior at larger plastic strains. This is corroborated by microtensile tests.

Accurate determination of frame compliance is an essential component of instrumented micro- and nanoindentation experiments. In load frames of finite stiffness, the load applied via the indenter induces displacement in both the sample and the load frame. Frame compliance must be identified and subtracted from the total indenter displacement to account properly for sample deformation. Current experimental procedures, in which the frame compliance is inferred from the elastic unloading indentation response of a reference sample, are based on several assumptions and simplifications that can propagate significant uncertainty with respect to subsequent analyses of mechanical behavior of the sample. We outline a new procedure that measures the compliance of the load frame, and identify the effects of load, loading direction, and loading rate on the measured compliance, as well as several possible sources of uncertainty in other available methods of frame compliance determination.

The morphology and distribution of high-pressure metastable phases BC8 and R8, formed in monocrystalline silicon under microindentation, were identified and assessed using transmission electron microscopy nanodiffraction analysis. It was discovered that the crystal growth inside the transformation zone was stress-dependent with large crystals in its central region. The crystal size could also be increased using higher maximum indentation loads. The BC8 and R8 phases distributed unevenly across the transformation zone, with BC8 crystals mainly in the center of the zone and smaller R8 fragments in the peripheral regions. Such phase distribution was in agreement with the theoretical residual stress analysis.

We used nanoindentation coupled with finite element modeling to determine the mechanical properties of amorphous Si layers formed by self-ion implantation of crystalline Si at approximately 100 K. When the effects of the harder substrate on the response of the layers to indentation were accounted for, the amorphous phase was found to have a Young’s modulus of 136 ± 9 GPa and a hardness of 10.9 ± 0.9 GPa, which were 19% and 10% lower than the corresponding values for crystalline Si. The hardness agrees well with the pressure known to induce a phase transition in amorphous Si to the denser β–Sn-type structure of Si. This transition controls the yielding of amorphous Si under compressive stress during indentation, just as it does in crystalline Si. After annealing 1 h at 500 °C to relax the amorphous structure, the corresponding values increase slightly to 146 ± 9 GPa and 11.6 ± 1.0 GPa. Because hardness and elastic modulus are only moderately reduced with respect to crystalline Si, amorphous Si may be a useful alternative material for components in Si-based microelectromechanical systems if other improved properties are needed, such as increased fracture toughness.

Nanoindentation was performed to evaluate the matrix strength of the ultra-fine grained steel produced by accumulative roll-bonding and subsequently annealed. The nanohardness, associated with the matrix strength, decreases with increasing annealing temperature. Because the matrix strength corresponds to the first term σ0 of the Hall–Petch relation, this result suggests that the σ0 might not be constant for these steels between various grain sizes. Therefore, it is suggested that the change of the macroscopic strength during annealing is dominated by not only the grain coarsening leading to a reduction of grain refinement strengthening, but also the softening of the matrix strength.

Electron beam physical-vapor-deposited Y2O3-stabilized ZrO2 thermal barrier coating (TBC) samples were indented from room temperature to 900 °C using an instrumented high-temperature vacuum displacement-sensitive indenter. Hardness and elastic modulus were determined from the load–displacement curves recorded during indentation. Both the hardness and the elastic modulus of the TBCs were much lower than those of dense ceramics of a similar composition; this is attributed to the increased compliance that results from the porous columnar microstructure of the TBCs. In addition, the TBCs showed an unusual absence of elastic recovery at the residual indents compared to the dense ceramics.