Multi-axis CNC machining and toolpath strategies for complex free-form geometries: a systematic review of geometry, kinematics, process physics and intelligent optimization

Abstract

In addition to offering the enabling process for the turbine blade, blisks, impellers, moulds, dies and other free-form components, multi-axis computer numerical control (CNC) machining introduces the geometry of toolpath, kinematics of machine, cutting mechanics, surface integrity and energy use. In this paper, research on five-axis and multi-axis toolpath strategies for complex geometries is summarized and critically analyzed, and the circumstances under which the various strategies are viable are identified. The literature was searched for in a systematic manner using the PRISMA 2020 reporting logic on studies published from 2016 to 2026. A total of fifty one primary studies and two previous reviews were coded into eight themes and summarized thematically at six interacting decision layers: geometry, toolpath generation, tool orientation, machine kinematics, process behaviour and optimisation. Fifteen research and development process optimisation, maintenance, Industry 4.0 and capability studies were conducted by the authors' research group and used to interpret the transferability into resource-constrained manufacturing contexts. There is no one dominant toolpath family throughout the categories of accuracy, cycle time, and stability and surface integrity. The most reliable techniques for controlling residual height are the "isoscalops" and "partitions" approaches, the "flank" approach and "trochoidal" or "barrel" (non-spherical tool) approach have advantages in terms of productivity or accessibility for appropriate geometry. On the part surface, an efficient path may not be efficient when generated via inverse kinematics, as singularities, rotary-axis reversals and feedrate limits produce a different realised path. Planning has been recently associated with models of cutting forces, digital twins, machine learning, reinforcement learning, but the connection of planning with industrial closed-loop validation, the generalisation across machines, and the consideration of energy and residual stress as planning goals are limited. The review describes a six-layer taxonomy and a coupled decision model for multi-axis planning, characterizes the corpus according to year and theme, and outlines a research agenda for the machine-aware, physics-informed and energy-aware computer-aided manufacturing (CAM) and the adoption of the technology in the development of industrial economies.