Nanoindentation testing has proved to be an effective tool to determine the mechanical properties of small volumes of
materials applied in various micro-systems, including hardness, indentation modulus, creep and so on. Nowadays, with
the help of advanced numerical methods, especially the finite element analysis (FEA) technique, further mechanical
properties of the material under test (e.g. tensile strength, etc.) can be interpreted from the typical indentation curve.
However, the reliability and accuracy of these analytical models have to be well tested.
Recently, the deformed topography of the interlayer surface within the tip-film-substrate system has been proposed to be
the reference for the evaluation of FEA and other mathematic models for indentation testing. Here an in-situ interlayer
deformation imaging system based on differential confocal microscopy is therefore developed, which has the capability
to measure in-situ the real-time topography deformation within a layered specimen during nanoindentation testing.
By means of linear regression and interpolation of the linear region of the standard confocal microscopy, differential
confocal microscopy (DCM) can achieve a very high resolution for topography measurements. However, the actual
capability and measurement uncertainty of DCM would be subject to those common-mode error sources like surface
heterogeneity, intensity fluctuation of the light source, etc. In this paper an improved DCM is proposed, which
introduces an additional point detector to the conventional DCM, creating dual confocal signals with slight relative axial
shifting. The real topography of the surface under test can then be easily deconvoluted from the dual differential signals,
whilst the common-mode errors within the measurement are eliminated.
A prototype was developed and applied for measuring a step-height composed of two different materials and for in-situ
inspection of the interlayer deformation during nanoindentation testing. Preliminary experimental results verify the
feasibility and accuracy of the proposed method.