How NASA Revealed HiddenProperty Variations with PIP Testing
Challenge
Where machining tensile specimens for testing is not an option, hardness is often considered as a solution.
This poses a significant limitation; while hardness testing offers advantages such as affordability, speed, ease of use, and suitability for testing small specimens, a hardness number is not a fundamental material property. Furthermore, practitioners are limited to conversions into a limited set of material property values, as hardness numbers cannot give full stress-strain curves, which means that without further information finite element modelling cannot always be conducted accurately.
Hardness numbers are produced by measuring the lateral extent or depth of an indent performed at a known load and applying a standard equation for the specific indenter used. These results are then frequently converted into UTS (ultimate tensile strength) and sometimes, albeit less frequently, yield stress values. Conversions can only be performed using empirical relationships, some of which can be found in standards published by ASTM and ISO.These standards contain warnings for the user, such as in BS EN ISO 18265:2013 where it states, âConversion from one hardness value to a tensile strength value involves uncertainties which must be taken into account.â
These uncertainties include the relationship between hardness and UTS, which is only available for certain specific categories of alloy, which limits the use of these procedures to these metals. Despite this hardness data is often used extensively to inform design/safety decisions and as a material quality release test. This reliance on hardness numbers alone poses a problem for failure analysis and end users alike as it can lead to the need for approximations as well as a loss of potential significant information about the materialsâ behaviour
Challenge
The aim of this case study, conducted in conjunction with HSE Science Division, was to determine whether PIP testing could provide the insights required for better understanding of the reasons for failure of components, as well as overcoming the challenge of testing of fragments from the failed component due to limited availability of the material.
In this study, we compare hardness, PIP (Profilometry-based Indentation Plastometry), and tensile testing to obtain mechanical properties and stress-strain relationships for a wide range of metals. Tensile testing is often considered the benchmark for determining these properties; however, it does come with some drawbacks. These include the size of sample required, the uniaxial nature of the results, and the need to machine specific geometries for testing. In comparison to this, hardness testing is much more convenient and can be performed quickly on a wide array of samples.
These advantages of hardness testing are shared by PIP. The difference between these techniques is in the results produced; with PIP producing a full stress-strain curve, similar to tensile testing, while hardness testing generates a single number for the material, from which other parameters can be determined based on empirical relationships.
The ability for mechanical characteristics to be extracted from this hardness number has been an area of substantial research, with many empirical relationships found and implemented in different situations.
Materials
Tensile specimens had 5 mm diameter and were tested with an extensometer gauge length of 25 mm. Indentation samples were in the form of small samples, with a thickness of at least about 5 mm and lateral dimensions of the order of 10-20 mm. Surfaces for indentation were polished to at least a 6 ”m finish.
12 different materials were used in this work:
- Waspaloy
- 316L Stainless Steel
- Rail Steel
- Aluminium 7075
- S355 Low Alloy steel
- Plain Carbon Steel
- Martensitic Steel
- 4340 Low Alloy Steel
- As-received Copper
- Carbon Steel
- 2D Steel
- Annealed Copper
Testing
Sample
NASA HR-1 (Fe-Ni superalloy) C-ring, manufactured by Laser Powder Bed Fusion (LPBF).
Measurements
Mechanical properties were measured using the PLX-Benchtop, a compact indentation-based device, equipped with a standard 1mm radius indenter
Method
Using PIP testing, Plastometrex mapped the bottom surface of the C-ring to create yield stress (YS) and ultimate tensile strength (UTS) maps
Measurements
The mechanical properties were measured using an Instron 3369 loading frame for uniaxial testing specimens and a PLX-Benchtop, a compact indentation-based device for PIP testing. PIP uses an accelerated inverse finite element method to infer accurate stress-strain curves from indentation test data.
The PLX-Benchtop comes with both 2 mm and 1 mm diameter indenter tips, allowing stress-strain measurements to be taken as close as 5 mm and 2.5 mm apart, respectively. The test itself is fully automated and takes less than 5 minutes without the need for separate samples or extensive sample preparation. The larger diameter indenter was used in this work for testing all samples.
Hardness testing was carried out using a Buehler Wilson VH3300 automatic Vickers hardness tester, with a load of 10 kg. The indent diagonal was taken to be the average of the two measurements (made via the built-in optical microscopy system).
In general, these two diagonals were very close, reflecting the isotropy of these materials, and also their relatively fine grain structures â even with the harder materials, the indents still tended to straddle at least several grains. These measurements were repeated in 5 different locations for each material.
Results
Yield stress fell ~15% (~90 MPa) as wall thickness decreased from 50 mm to 10 mm, while ultimate tensile strength stayed largely constant.
Average PIP results showed strong agreement with NASAâs independent tensile data: YS was within 2.6% and UTS was within 0.4%, confirming the accuracy of the test.
.webp)
Figure 2: Plot of (a) yield strength and (b) ultimate tensile strength as a function of c-ring thickness showing significant increases in yield strength and minor decreases in UTS as c-ring thickness increases
Figure 2: Plot of (a) yield strength and (b) ultimate tensile strength as a function of c-ring thickness showing significant increases in yield strength and minor decreases in UTS as c-ring thickness increases
Results
In this case study, results from uniaxial tensile testing were compared with those obtained from PIP tests and with hardness tests converted to UTS values.
PIP testing provides the full stress-strain curve of each material directly as part of its data output. However, hardness numbers cannot be directly compared to full tensile tests, as empirical relationships are used to convert hardness values to UTS values.
One example of these empirical relationships is found in BS EN ISO 18265:2013 for unalloyed and low alloy steels and cast steels. It is important to note the text of this standard, which describes the caution that should be used when applying this relationship:
âWhile hardness-to-hardness conversion involves considerable scatter and systematic errors, conversion of hardness to tensile strength values produces even greater scatter. One reason for this is that a great uncertainty can be affected by microstructural changes (e.g. resulting from heat treatment or cold working) within even the same type of steel.â
This is described in the standard as scatter bands of ±25 Hv10 for the hardness and ±85 MPa for the UTS.
In addition, this standard method can only be used on specific subsets of alloys, such as âunalloyed and low alloy steelsâ, whereas PIP can be applied to any metallic material and can account for the effects of microstructural changes which result in different mechanical responses of the material.
.jpg)
Figure 2: Nominal stressânominal strain curves for selected materials, from tensile testing and from PIP. The latter are shown only up to the onset of necking (peak in the plot).
This standard method was applied to the subset of materials in this study for which the conversion is permitted. PIP tests were carried out on all the samples, and the UTS values obtained from these two methods are compared on a unity plot with the tensile measurements in Figure 1. Perfect agreement with tensile testing is indicated by the solid unity line, with deviations of ±10% shown by dashed lines.
PIP values of UTS align 2.96 percentage points more closely with tensile measured values than hardness converted values, having a mean absolute percentage error of 3.8% compared to 6.8% for hardness testing.
While hardness conversions can be used only to estimate UTS values, PIP can also provide the full stress-strain curve, including yield stress and hardening behaviour. Figure 2 demonstrates the agreement between the full stress-strain curves from tensile testing and PIP for a subset of the materials in this work, covering a wide variety of properties.
In general, the level of agreement between PIP-derived and directly measured tensile curves is very good. This full stress-strain curve cannot be obtained from a hardness test, which can result in blind spots when trying to use hardness values to understand the mechanical properties of metallic components.
Conclusion
As demonstrated, hardness testing within failure analysis may be used as a complementary technique to tensile testing. When use of tensile testing is not possible, hardness numbers can provide useful indicators of the resistance to plastic deformation.
Empirical correlations from standards can be used to convert these hardness numbers to UTS values, for a limited subset of alloys. These standards cannot be applied universally and detail that conversions should âbe used with cautionâ.
PIP testing revealed variations in strength that would have been missed with conventional methods, giving NASA the data to connect local performance to build conditions and geometry. With this information, future manufacturing runs can incorporate adjusted print parameters or tailored geometries to maintain structural integrity without compromising safety.
