User talk:BillClyne/sandbox

Indentation Plastometry
I think that there may be benefits in creating a new article entitled "Indentation Plastometry". There is great interest in obtaining mechanical properties (of metals) via indentation procedures. This dates back over 100 years, when the first Hardness Tests were introduced. There is an existing article on "Hardness" (https://en.wikipedia.org/wiki/Hardness) and also one on "Indentation hardness" (https://en.wikipedia.org/wiki/Indentation_hardness), which is the main way of obtaining a hardness number. The problem with hardness numbers is that they have only a loose and ill-defined relationship with the real mechanical (plasticity) characteristics, which are fully captured by the (true) stress v. (true plastic) strain curve. Such curves are conventionally obtained by uniaxial testing - usually in tension, although compressive testing is also possible. These curves are needed for any quantitative assessment of how a particular metal component will perform in service, often carried out using the "Finite element method" (https://en.wikipedia.org/wiki/Finite_element_method). However, "Tensile testing" (https://en.wikipedia.org/wiki/Tensile_testing) is much more cumbersome and difficult than indentation testing, which also has the advantages of being non-destructive and allowing the mapping of properties over a surface, or monitoring changes with depth below the surface.

There have, however, been recent developments that allow reliable stress-strain curves to be obtained using indentation-based techniques, provided certain conditions are met. These procedures are now often grouped under the term "Indentation Plastometry". They have evolved partly because indentation set-ups have become more advanced, and they now commonly allow both acquisition of load-displacement data during the test and/or accurate measurement of the residual indent profile afterwards. There are several ways in which such data can be converted to stress-strain curves. The simplest is direct conversion of a load-displacement curve (using approximations that relate load to stress and displacement to strain), although it tends to be rather inaccurate. More complex, but also with much greater potential for accuracy, are procedures in which the test is repeatedly simulated using the finite element method, changing the stress-strain curve used as an input until optimum agreement is reached between measured and modelled outcomes. There are also important issues relating to the scale of the indentation and to the plastic strains induced in the sample. In particular, much "Nanoindentation" (https://en.wikipedia.org/wiki/Nanoindentation) involves deformation of a volume that is too small to be representative of the response of the bulk. This depends on the scale of the microstructure of the metal. Similarly, very shallow indentation may induce only small plastic strains, such that the test outcome cannot be used to capture the stress-strain curve over an appropriate range of strain (typically up to several tens of %). There are also important points to note regarding the effects of inhomogeneities in the sample and the possibility that it exhibits "Anisotropy" (https://en.wikipedia.org/wiki/Anisotropy) - ie different stress-strain curves when measured in different directions.

There have been many publications over the past couple of decades that are relevant to this topic, and many claims made. It's an area that is potentially very confusing for the newcomer. However, certain points have become clear recently and I think that it would now be timely to attempt to distil a short article from this plethora of information, citing a suitable number (probably ~ 30) of these papers. I have a rough draft of such an article, potentially with a few illustrative figures. Before attempting to upload any of this material, I'd appreciate hearing any views on the viability of such an article. Of course, there may be a substantial number of people who would subsequently be interested in editing it. BillClyne (talk) 14:47, 3 August 2022 (UTC)