Calibration And Uncertainty Evaluation Of Nanoindentation Instrument

Nanoindentation measurement technique has many advantages，such asflexible operation, high measurement resolution, nondestructive or slightlydamaged to samples, etc, which has been widely applied in material mechanicaltesting of metals, micro-electromechanical systems, composites, biomaterials,films/coatings and so on. In the commercial nanoindentation instrument,measurement errors can be due to the sensors of displacement and force.Moreover, the morphology of indenter has certain discrepancy with thetheoretical model. These errors will decrease the measurement accuracy.Therefore, it is very necessary to calibrate the nanoindentation instrument.Based on the previous calibration work of displacement scale factor, theindenter area function and load scale factor of nanoindentation instrument werecalibrated. Based on the above calibration works, the measurement uncertaintyof nano-hardness and reduced modulus of the equipment was evaluated in thisthesis. The main research works and relative results are listed as follows:1. Calibration and traceability of the Berkovich indenter area function. Themorphology of indenter has the deviation to the geometry of theoretical model,which is mainly due to the usage wear and manufacturing precision. Two methods were adopted to calibrate the area function of Berkovich indenter: onemethod is the direct calibration method based on measurement of atomic forcemicroscope (AFM), in which three kinds of scanning frequency (0.2Hz,0.5Hz,1Hz) were used for the scanning of Berkovich indenter. The indenter areafunction was subsequently obtained by fitting the relationship betweencross-sectional areas and the heights of the indenter. The other method is aindirect way, which is based on the indentation experiment of the standard fusedsilica with the known elastic modulus. With the calibrated value of the machinecompliance, a series of indentations of different depths were applied on thefused silica and the indenter area function was then determined by a iterativemethod. The calibrated indenter area functions by both methods were applied tothe indentation tests of the standard samples of fused silica and polycarbonate.The indentation results show, at low indentation depth, the hardness values andthe reduced modulus values measured by the AFM method of0.5Hz are moreaccurate than that of values calibrated by the indirect method, however, theresults of the two methods have small difference at large indentation depth.2. Calibration and traceability of the load scale factor. It was found that inthe process which using the traceable standard weights to calibrate the load scalefactor， the load condition of indenter and specimen didn’t get completelyreappeared. Therefore, the force standard device of type FS-C130was used tocalibrate the load scale factor of nanoindentation instrument. According to theresults of calibration experiments, the average of load scale factor was 1.0131mV/mg, the standard deviation was0.0003mV/mg. The calibrationresults of the indenter area function and the load scale factor were applied to thesystem of nanoindentation instrument. A series test of different indentationdepths were performed on the standard fused silica and the measurement resultswere compared with that of values before the calibrations. The comparisonresult show, after the calibrations, the measurement values are more close to thenominal values of the standard samples than that of values before thecalibrations, which indicates the measurement accuracy of the equipment wasremarked improved after the calibration works.3. Uncertainty evaluation of nanoindentation instrument. The mainmeasurement uncertainty factors were load, displacement, indenter area functionand contact stiffness. Then the uncertainty of each component of thesefactors (such as contact zero point, thermal drift, machine compliance, noise, etc)was calculated respectively. The uncertainties of nano-hardness and reducedmodulus were finally evaluated by the combination of the those uncertaintycomponents. The results are that the expanded relative standard uncertainty ofnano-hardness and reduced modulus is5.86%and2.94%, respectively, both at aconfidence level of95%.