Drilling thrust force in a SiO₂-reinforced AZ91 magnesium composite rose from 118 N to 1,538 N across a 30-to-110 m/min speed range, and torque climbed from 0.23 N·m to 8.23 N·m, according to an experimental study published in Nature.
The extreme runs differed in more than speed: the lowest-load run paired 30 m/min with a 0.05 mm/rev feed, vegetable oil and a high-speed steel tool, while the highest-load run combined 110 m/min, 0.15 mm/rev feed, liquid nitrogen cooling and uncoated carbide.
The study, an experimental investigation and multi-criteria assessment of drilling the composite under bio-based and cryogenic cooling conditions, also recorded surface roughness values spanning 0.13 μm to 4.08 μm. The authors warn that the observed changes in thrust force, torque and roughness represent the combined outcome of the full drilling setup and cannot by themselves be taken as proof of any particular mechanism of material removal.
Speed, feed and the composite itself all appear to shape the loads. According to the paper, a higher feed can enlarge the uncut chip thickness, deepen penetration resistance and enlarge the amount of material the cutting edges must engage, whereas a higher cutting speed can change interfacial velocity, frictional dissipation, strain rate and the thermal condition of the AZ91 matrix.
The stiff SiO₂ particles may additionally influence these behaviors by resisting matrix deformation locally and by interacting with the tool at intervals.
A descriptive marginal analysis averaged across six observations at each cutting-speed level showed mean torque rising from 0.475 N·m at 30 m/min to 5.012 N·m at 110 m/min, mean roughness from 0.240 to 1.893 μm, and mean thrust force from 164.3 to 1059.7 N. Across feed levels, the averages moved from 2.070 to 2.987 N·m, from 0.688 to 1.147 μm, and from 492.0 to 657.1 N as feed rose from 0.05 to 0.15 mm/rev.
The response gradients associated with cutting speed were larger than those associated with feed rate over the ranges investigated.
The authors avoid claiming a hierarchy of factor dominance. Because coolant and tool levels were not independently crossed with every speed-feed condition, coolant- and tool-related trends are treated as conditional observations, and direct causal attribution to either categorical factor is avoided.
Post-drilling scanning electron microscopy of the burr retained from the original lowest-load specimen provided direct morphological evidence of localized microcracking and matrix tearing within that burr — observations the study restricts to that examined burr and does not generalize to the other drilling conditions.