It is apparent that increasingly testing of disc shaped samples using a ball or hemispherical indenter is being advocated for measuring materials properties. It should be apparent that at the start of loading the disc must deform to take on the shape of the ball. This is a relatively high deformation process. For very brittle samples there is insufficient ductility available and cracking occurs. For ductile materials there is the risk of the strain introduced modifying the properties. Thus, the structure when the measurement is taken is significantly altered. In both cases the results obtained may not be representative
From my experience, there is some great work done in the past by Thomas Hyde and recently by J Rouse at Nottingham University on small punch creep testing specifically. This is my first go to on such testing/modelling. You are absolutely right that the punch will affect the local response differently depending on the underlying sampled material characteristics and its strain rate sensitivity. It would be interesting to employ advanced modelling such as physically based CPFE models and in-situ characterisation techniques in parallel to explore this on common engineering alloys, say Grade91, 316/304 and the Inco family, but also more advanced and novel alloys with potential future application and code qualification.
Thanks Mark, I think one key is that the test method may not develop representative data for all materials and test geometries. Thus, is is important to undertake appropriate checks and validation
I completely agree with Jonathan David Parker and Mark Petkov . This question should interest the Project
https://www.researchgate.net/project/Standardization-of-the-Small-Punch-test-tensile-and-creep
Personally, I think that small punch tests should never replace the uniaxial test. Nevertheless, there would be a role for small punch tests when (1) material is limited and (2) the problem being investigated requires large strains.
The "large strain" issue is the key point to Jon's question.
Like all the other small sample testing techniques, a small punch test may not be representative (especially when the grain size of a material is large). Small punch can also be used as a rapid mechanical characterisation (pre-screening) method for AMed alloys. Perhaps, the small punch test cannot replace the uniaxial test, however, a correlation relationship (phenomenological or empirical) between the uniaxial test and the small punch test can be always established.
Please be very careful the fact that for some metals and alloys an empirical correlation can be shown for a specific range of conditions should NOT be taken as evidence that the results agree over all conditions and for all alloys. This emphasis is simply to ensure that details are properly considered. I will post some links to relevant papers if that would be helpful
As with most small-scale test technologies (SSTT), I think we should appreciate the small punch test (SPT) for what it is. The SPT has been around for decades and originates from nuclear research where irradiation volume is scarce and in the semi-non-destructive properties testing with scoops from existing installations. It can correlate test data with known large-scale testing (LST) data with well-defined stress states. SSTT can be useful for tracking properties development and, if the constitutive equations are known with high confidence, modeling may allow estimation of the response of a larger scale component.
The above being said, it is vital to recognize the influence of the stress state during SPT on the mechanical response and be aware of the limitations of continuum mechanics modeling. Continuum models are approximations and generalizations from large scale tests. Models may over-simplify, or not take account of, effects taking place only in large scale testing (such as grain size variation over the thickness or the surface state of a forging). Conversely, thin specimen failure modes may not be representative of large-scale component response. With the complicated development of the stress state during SPT and the absence of 3D stress in such thin wall specimens, estimation of material properties without LST calibration is likely to be inaccurate.
For instance, testing the ductile to brittle transition (in terms of fracture toughness or impact DBTT) of ferritic steels has proven to be inaccurate. But the SPT may give useful results for measuring "bulk" properties such as yield strength or creep rates. But, in my opinion, it cannot replace standardized large-scale testing for e.g., code data purposes.
- Badges
- Science topic
- Similar topics
- Engineering
- Mechanical Engineering