Depth-sensing indentation test

Depth-sensing indentation tests were conducted using a Micro Materials Ltd.

NanoTest^TM NTX equipped with a Berkovich indenter. The NanoTest^TM head has a load range of 0-500 mN and maximum load and depth resolutions better than 100 nN and 0.1 nm, respectively. A new Berkovich diamond indenter with a tip radius of less than 20 nm, shown in Figure 3.3, has been used to perform the indentations. The properties of the indenter were taken as E = 1000 GPa and υ = 0.07 for calibration and analysis [14].

Figure 3.3. SEM images of the Berkovich indenter tip using secondary electrons at an acceleration voltage of 10.0 kV.

Sample fixtures may add to the compliance of the instrument and thus, cyanoacrylate adhesive has been used to glue the mounted sample to the instrument holder. Before testing, both the indenter area function and the load frame compliance were calibrated using a standard sample of fused silica (E = 72 GPa, v = 0.17 and H = 9 GPa [123]) in compliance with ISO 14577 [13]. In order to remove the effects of surface roughness

Chapter 3- Research methodology

39 on the results, indentation loads were selected to reach maximum indentation depths of at least 20 times the average roughness (Ra) of the specimen in accordance to ISO14577 [13]. Ra is defined by ISO 4287/1-1997 [124] as the arithmetic average surface height deviations measured from the mean plane:

𝑅_𝑎=1 𝐿∑ 𝑍_𝑗

𝐿 𝑗=1

3.1

where Zj is the profile height function analysed in terms of height (Z) and position (j) over the evaluation length L. Given that the polishing procedure affects the surface of the specimen to a depth of about the same size as the nominal grit due to strain-hardening or cold-working [14], indentations were performed at sufficient load to ensure indentation depths of at least three times the thickness of the strain-hardened layer as suggested by Liu et al. [36]. Therefore, the indenter has been loaded from an initial contact force (Pi) of 0.1 mN to a maximum force (Pmax) within the range of 30-480 mN at a loading and unloading rate of 10 mN/s for CrMoV and Ti-6Al-4V. C110 on the other hand was loaded and unloaded at a rate of 4 mN/s in order to obtain a similar number of data points, considering that this material is indented with less than half the load of all other materials. In specimens of CrMoV, a dwell period of 30 s at Pmax was applied so as to ensure the unloading data used for analysis purpose were mostly elastic. In addition, the load was held constant for a period of 30 s at 0.1Pmax to establish the rate of displacement produced by thermal expansion in the system, that is, thermal drift. Therefore, thermal drift corrections were performed in addition to these curves. However, due to the complications in the inverse analysis caused by a non-continuous P-h curve, this load-time sequence was avoided in C110 and Ti-6Al-4V and instead, a single loading and unloading ramp was defined. In order to assess the sensitivity of the P-h curves to the loading-unloading rate, an extra set of indentations in titanium at a rate of 1.5 mN/s have been performed. Sets of five indentations were performed per indentation test, at an offset of at least 20 times the maximum depth as suggested by ASM International [123], in order to avoid overlapping of plastic strain-hardened zones. The possible sensitivity of the depth-sensing indentation data to the thickness of the mechanically-hardened layer due to

Chapter 3- Research methodology

40 polishing has been assessed by polishing an additional sample of CrMoV to 1 µm using a diamond suspension. This material was selected for assessment as the fine microstructure, compared to those of the Ti-6Al-4V and C110, is expected to affect to a lower extent the experimental data. Table 3.3 provides details of the parameters used in each set of indentations.

With the intention of having an equivalent level of uncertainty in the extracted data, the P-h curves reaching similar indentation depths have been selected to recover the properties of the material via the inverse analysis of indentation. Therefore, as indicated in Table 3.3, Pmax = 240 mN for Ti-6Al-4V and CrMoV, and Pmax = 120 mN for C110. Those identified with an asterisk have been used in the inverse analysis procedure as detailed in the following section.

Table 3.3. Parameters used in the different depth-sensing indentation tests.

Specimen Ti-6Al-4V CrMoV C110

L-U rate [mN/s] 10 1.5 10 4

* Indentation parameters of the P-h curves used in the inverse analysis.

Building upon the experience gained from the analysis of the depth-sensing indentation data extracted from the three single materials, the indentation experiment for the SCMV-SCMV IFW was set as follows: indentations were performed using a single loading and unloading ramp at 10 mN/s, i.e. no dwell period at Pmax, from a Pi

= 0.02 mN to a Pmax = 240 mN. The system was left over night to thermally stabilise

Chapter 3- Research methodology

41 to reduce thermal drift effects during measurements. A dwell period of 20 s at 90%

unload was defined for drift correction in the indentation schedule to confirm the negligible effects of heat transfer; this time is approximately the same as the time for application or removal of the test force. The indentation unit has been setup to conduct five rows of indentations, with an offset of 50 µm, along the joint at distances of 0, 0.5, 1.5, 3, 4.25, 4.5, 4.625, 4.75, 5.25, and 8 mm from the weld line as illustrated in Figure 3.4. Therefore, depth-sensing indentations are aligned with microhardness impressions, to at least 250 µm above the microindentations scan line so as to test strain free volumes of material.

Figure 3.4. Schematic diagram of indentation sites across the IFW joint. Units are in mm unless otherwise specified. The diagram is for illustration purposes only and thus not in

scale.