This paper, based on extensive finite element simulations and scaling analysis, presents scaling functions for the inverse problem in nanoindentation with sharp indenters to determine material properties from nanoindentation response. All the inverse scaling functions were directly compared with results calculated using the large deformation finite element method and are valid from the elastic to the full plastic regimes. To relate the material properties to measurable indentation parameters a new nondimensional experimental parameter Λ=P/(DS) was introduced, where P is load, D is indentation depth, and S is contact stiffness. This parameter is monotonically related to the ratio of yield stress to modulus. The modulus, hardness and yield stress are presented as explicit functions of Λ and the strain hardening exponent. The error in the inverse modulus, hardness, and yield stress due to uncertainty of the strain hardening exponent was studied and is compared with that of the traditional Oliver–Pharr method. The method of determining the strain hardening exponent from measurement with an additional indenter with a different cone apex angle is described. For this, a scaling function with the strain hardening exponent as the only unknown was obtained. In this way, the modulus, hardness, yield stress and strain hardening exponent may be determined. Experimental results show the inversion method permits the modulus and hardness to be accurately determined irrespective of the effects of pileup or sink-in.

No abstract

Using an electron beam melting (EBM) printing machine (Arcam A2X, Sweden), a matrix of 225 samples (15 rows and 15 columns) of Ti-6Al-4V was produced. The density of the specimens across the tray in the as-built condition was approximately 99.9% of the theoretical density of the alloy, ρT. Tensile strength, tensile elongation, and fatigue life were studied for the as-built samples. Location dependency of the mechanical properties along the build area was observed. Hot isostatic pressing (HIP) slightly increased the density to 99.99% of ρT but drastically improved the fatigue endurance and tensile elongation, probably due to the reduction in the size and the distribution of flaws. The microstructure of the as-built samples contained various defects (e.g., lack of fusion, porosity) that were not observed in the HIP-ed samples. HIP also reduced some of the location related variation in the mechanical properties values, observed in the as-printed condition.

Carbon-doped oxide SiCOH films with low to ultralow dielectric constants were prepared on a Si substrate by plasma-enhanced chemical vapor deposition (PECVD) from mixtures of SiCOH precursors with organic materials. The films have different levels of nanoscale porosity resulting in different dielectric constants and mechanical properties. The mechanical properties of the films have been characterized by continuous-stiffness nanoindentation measurements. To study the effect of film thickness, each group of samples with the same dielectric constant was composed of samples prepared with different film thicknesses. It is shown that the effective hardness and modulus of the SiCOH/Si substrate system depends significantly on indentation depth due to substrate constraint effects. The "true" film properties were determined using both an empirical formulation of the effective modulus and direct inversion based on a finite element model. The hardness and modulus of three groups of samples with different degrees of dielectric constants have been measured. The hardness increases from 0.7 to 2.7 GPa and modulus from 3.6 to 17.0 GPa as the dielectric constants change from 2.4 to 3.0. While for stiffer films the modulus measured at an indentation depth 10% of the film thickness is close to the "true" value for films thicker than 0.5 m, the measured value can give an overestimate of up to 35% for softer films. Thin film cracking and film-substrate debonding have been observed with scanning electron and atomic force microscopy at the indentation sites in softer films. The damage initiation is indicated by pop-in events in the loading curve and sharp peaks in the normalized contact stiffness curves versus indentation depth.