Effects of Feed Rate and LN₂-Derived Cryogenic N₂ Gas-Jet Cooling on the Surface-Altered Layer of TC18 Titanium Alloy during Turning

Auxiliary cryogenic-only series on cutting speed and depth of cut are also included.The dataset comprises four categories of core data files. The first category consists of material property tables (Tables 1 and 2), recording the nominal chemical composition of TC18 titanium alloy (Ti as balance, Al 5.28%, Mo 5.12%, V 5.15%, Cr 1.12%, Fe 0.11%, Zr 0.01%, C 0.01%, N 0.00%, H 0.00%, O 0.11%) and mechanical property parameters (ultimate tensile strength 1198 MPa, 0.2% proof stress Rp0.2 1128 MPa, elongation at break 14.5%, reduction of area 52%), as certified by the material certificate (No. 20250268, Shaanxi Huachen Titanium Co., Ltd.).

The second category is the experimental conditions table (Table 3), detailing all process parameters including cutting speed, feed rate, depth of cut, cooling medium, delivery method, nozzle parameters, pressure, and temperature measurement method. The third category is the surface integrity summary table (Table 4), containing near-surface microhardness (HV), hardened layer thickness (μm), plastically deformed zone (P-zone) thickness (μm), and surface roughness Ra (μm) for all eight cutting conditions, with the number of measurement positions (n = 5) indicated for each metric and data presented as mean ± standard deviation.

The fourth category comprises raw measurement data, including microhardness profile data along depth for each condition (measured from approximately 10 μm below the machined surface at 10 μm intervals, with 5 μm refinement in the transition zone; five independent indentations laterally offset by approximately 30 μm at each target depth), five replicate raw surface roughness measurements per specimen, and metallographic micrographs (cross-sectional morphology of the near-surface plastically deformed zone captured at 500× optical magnification).The data acquisition and processing procedures are as follows.

After turning, specimens were sectioned into 10 mm × 10 mm × 5 mm metallographic samples using a TRS-32 wire electrical discharge machine, hot-mounted, ground, polished, and etched with Kroll's reagent (HF 2 vol% + HNO₃ 4 vol% + deionised water 94 vol%). The plastically deformed zone (P-zone) was observed using an Ossila M230-3M180 metallographic microscope equipped with a 20-megapixel USB3.0 camera (chip size 1/1.8″, 7.17 × 5.38 mm); images were acquired at 500× optical magnification and measured using Image-Pro software, with pixel-to-length calibration performed using a stage micrometre prior to measurement.

For each cutting condition, at least three representative fields of view were examined on a single cross-section; the P-zone boundary was delineated independently by two trained operators, and where discrepancies exceeded 5%, the images were re-examined jointly to reach consensus, with reported values representing the mean of repeated measurements. Near-surface microhardness was measured using an HVS-1000Z Vickers microhardness tester at a test load of 50 gf with a dwell time of 15 s.

The matrix-hardness reference band (HV 376–388, nominal central value HV 384) was established from 25 measurements (five indentations at each of five depths separated by 20 μm) in the stable region beyond 200 μm from the machined surface, defined as the mean ± 2 standard deviations. Hardened layer thickness was defined as the first depth at which the hardness profile returned to and remained within the matrix-hardness band; given the 10 μm measurement step, its uncertainty was conservatively estimated as ±10 μm.

Surface roughness was measured using an OPTELICS HYBRID confocal microscope (Lasertec Corporation, 405 nm violet semiconductor laser source, vertical measurement accuracy 0.05 nm, horizontal repeatability 3σ < 10 nm) with an evaluation length of 4.0 mm and a cutoff wavelength of 0.8 mm (Gaussian filter) in accordance with ISO 4287; measurements were performed perpendicular to the feed marks, and five distinct positions were examined on the machined surface of each specimen, with mean and standard deviation calculated from these five measurements.

Data plotting and statistical processing were performed using Origin software, and metallographic image post-processing used Photoshop and AOSVI software.The following notes on data quality and limitations apply. In this study, only one specimen was machined and one cross-section prepared per cutting condition. Among the reported mean ± standard deviation values, the microhardness standard deviation reflects spatial variation from five independent indentations at the same depth on a single cross-section, while the surface roughness standard deviation reflects measurement repeatability across five positions on a single specimen; neither represents variation across independently machined specimens.

Accordingly, all comparisons between cooling modes are descriptive observations, and statistical significance cannot be assessed in the absence of independent machining replication. The first near-surface indentation centre was positioned approximately 10 μm below the machined surface; at a load of 50 gf and hardness range of HV 376–446, the estimated indentation diagonal is approximately 14.4–15.7 μm, which does not satisfy the ISO 6507-1 requirement that the distance from the indentation centre to the specimen edge be at least 2.5 times the mean diagonal length (approximately 38 μm).

The near-surface hardness values should therefore be interpreted as qualitative indicators of the hardness trend rather than absolute measurements. P-zone depth measurements rely on manual boundary delineation from 500× optical metallography; although independently measured by two operators, formal blinded repeatability assessment was not documented, and measurement uncertainty was estimated from the standard deviation of boundary placements across multiple fields, with reported values rounded to appropriate decimal places to reflect actual measurement precision.

Hardened layer thickness values are quantised by the 10 μm measurement step (20, 30, 40, 50, 60 μm, etc.), and differences between cooling modes of only 10 μm (i.e., one measurement interval) should be interpreted as trends rather than precisely quantified effects. No missing values exist in the dataset; all four metrics were measured for all eight main experimental conditions. All tabular data files are in .xlsx format and can be opened with common spreadsheet software such as Microsoft Excel, WPS Spreadsheets, or LibreOffice Calc; metallographic images are in standard .jpg/.png formats viewable with any image viewer.