Statistical correlation between microstructural features and process forces in conventional and ultrasonic-assisted milling of plasma claddings

Wear-resistant NiCrSiFeB hardfacings protect highly loaded components but are difficult to finish-machine because of their high hardness and heterogeneous hard-phase population. Previous studies primarily compared individual alloy variants, while quantitative relationships between hard-phase morphology, cutting-force evolution and tool wear remain insufficiently understood. This study establishes an exploratory descriptor-based correlation between the microstructure and machinability of modified Colmonoy 56 hardfacings. Minor additions of Al, Ti, Nb, Hf and Mo were investigated using cast screening specimens, and selected modifications were transferred to plasma-transferred-arc claddings for process validation and wear assessment. Hard-phase fraction, number, size, shape, distribution and statistical dispersion were quantified and related to cutting forces, cutting-force slope, flank wear and surface roughness during conventional and ultrasonic-assisted milling. The alloying additions substantially altered precipitation morphology, hardness and machinability. Al and Ti promoted elongated and branched precipitates associated with increased cutting forces and tool wear, whereas Nb produced the most favourable combined machining response. Dispersion measures describing hard-phase size, shape and distribution indicated additional microstructure-sensitive associations with the cutting response beyond those observed for mean values alone. Moreover, the cutting-force slope over removed material volume correlated strongly with tool flank wear, while the initial cutting force was not a reliable wear predictor. The favourable behaviour of the Nb-modified alloy was confirmed in plasma claddings, where improved machinability was achieved while maintaining wear resistance close to that of the reference alloy. These findings support the consideration of hard-phase heterogeneity in addition to conventional mean descriptors and provide hypotheses for the microstructure-oriented design of wear-resistant claddings with improved finish-machining performance.