Improvement of predicting mechanical properties from spherical indentation test

Although nowadays the Instrumented Indentation Test (IIT) can be applied to the fields from nano-micro- to macroscopic scale, this paper only addresses its application in the area of life assessment of components in industry, such as power and/or petroleum chemistry plants. In comparison of IIT with other miniature specimen techniques, such as the small punch test, impression test, micro-tensile or small ring tests etc., the relative advantages of the IIT technique are: non-destructive test could be performed on site, no sample machine is needed to cut out material, thus no sample processing is necessary. Therefore, the IIT method has potential for residual life assessment of components in service. A portable IIT indenter has been developed to measure both load and displacement, from which actual material properties of components can be evaluated. In order to develop analytical software for a new portable IIT instrument, authors reviewed several existing analytical methods, such as the representative stress-strain method, dimensional analysis method, and inverse finite element method. This paper first gives an overview of these analytical methods, their advantage and disadvantage, and then put forward ideas to improve them. Finally, the so-called Neural Network (NN) is introduced as the NN method. This method, can deliver evaluation on site without time-consuming finite element analysis. Experimental results for IIT are provided by DNV KEMA. First verifications are carried out for elastic plastic properties, such as yield strength, tensile strength, Young's modulus and Brinell hardness. A good agreement is found between the conventional test results and the analytical prediction from IIT results. More validation and tests are needed and planned for the near future. (C) 2016 Elsevier Ltd. All rights reserved.