Improving the accuracy of nanoindentation instruments requires multidimensional control from instrument calibration, sample pretreatment, parameter optimization, and model adaptation, combined with repeated experiments and cross method validation, in order to obtain reliable micro nano mechanical data. Nanoindentation technology, as the core means of characterizing the micro nano scale mechanical properties of materials, directly affects the evaluation of key parameters such as hardness and modulus of materials based on the accuracy of its test results. However, the coupling effect of multiple factors during the testing process may lead to data bias, which needs to be systematically analyzed from three aspects: instruments, samples, and operations.
1、 Instrument hardware and calibration accuracy
The performance of the core components of the nanoindentation instrument, such as the indenter, displacement sensor, and load sensor, directly determines the measurement limit and resolution. If there is wear or machining error in the geometric shape of diamond indenters (such as Berkovich pyramids), it will change the calculation benchmark of contact area, resulting in systematic deviation of hardness and modulus. In addition, the calibration status of sensors is crucial: load/displacement sensors that are not regularly calibrated can introduce zero drift or sensitivity deviation, such as load sensors not resetting after overload, which may distort data in low load areas. Although high vacuum environment can reduce air damping interference, fluctuations in environmental temperature and humidity may still affect sensor stability, and it needs to be operated in a constant temperature and humidity laboratory.
2、 Sample characteristics and surface state
The surface roughness of the sample is the primary interfering factor. If the surface roughness exceeds 1/10 of the indentation depth (such as Ra>50nm), the indenter is prone to contact the convex peak first rather than the plane, causing "false indentation" and resulting in an artificially high measured modulus. For thin film samples, the substrate effect is significant: when the film thickness is less than 10 times the depth of indentation, substrate deformation will be superimposed on the film response, and it needs to be corrected through the Oliver Pharr model or adopt ultra shallow indentation mode (depth<100nm). In addition, the anisotropy of the sample (such as different crystal planes of a single crystal) can lead to changes in mechanical properties with the direction of indentation, and the testing orientation needs to be clearly defined; Porous or composite materials may cause cracks due to local stress concentration, resulting in abnormal load displacement curves.
3、 Test parameters and operating procedures
The loading rate and holding time affect the rate dependent response of the material. The modulus of viscoelastic materials such as polymers is sensitive to loading rate, and excessive loading (such as>10mN/s) may mask creep behavior, leading to overestimation of modulus. If the time during the loading stage is insufficient (such as<10s), the viscoelastic relaxation is not completed, and the unloading curve deviates from the ideal elastic line, which affects the calculation of contact stiffness. The selection of indentation depth needs to balance resolution and substrate interference: too shallow (<20nm) is easily affected by surface oxide or adsorption layers, while too deep may penetrate the thin film. In addition, alignment deviation between the indenter and the sample (such as tilt>2 °) can lead to incorrect calculation of contact area, and it is necessary to ensure vertical pressing through optical or laser positioning.
4、 Applicability of data processing models
The classic Oliver Pharr model assumes that the material is a homogeneous and isotropic elastic body, ignoring the nonlinearity of stress distribution in the plastic zone. For superhard materials such as diamond-like carbon films or large plastic deformation scenarios, this model may underestimate the true modulus; However, the viscoelastic properties of soft materials (such as biological gel) need to be modified by introducing Kelvin Voigt model. In addition, the calculation of contact depth depends on the area function of the indenter. If it is not calibrated by a standard block (such as fused silica), the small error of the area function will be amplified to the modulus result (the error transfer coefficient is about 2).