Preheating and thermal behaviour of a rotating cylindrical workpiece in laser-assisted machining

Abstract

Laser-assisted machining is a widely used technique for preheating workpiece to reduce cutting forces and promote machinability in metal machining, thereby enhancing manufacturing quality and productivity. In setting laser-assisted machining parameters, the current practice typically relies on trial-and-error approaches. The uncertainties thereof could lead to adverse outcomes in product manufacturing, thus negating the potential benefits of this machining method. A clear understanding of workpiece thermal behaviour under laser spot heating is pivotal to developing a systematic basis for determining required preheating levels and optimised cutting variables for laser-assisted machining. In achieving this, the experimental methods are recognised to be largely impractical, if not tedious, due to instrument limitations and practicality of suitable non-intrusive measuring methods. Conversely, numerical methodologies do provide precise, flexible and cost-effective analytical options, warranting potential for insightful understanding on the transient thermal impact from laser preheating on rotating workpiece. Presenting such an investigation, this article presents a finite volume-based numerical simulation that examines and analyses the thermal response imparted by laser spot preheating on a rotating cylinder surface. On a rotating frame of reference using the ANSYS Fluent solver, the numerical model is formulated, accounting for transient heat conduction into the cylinder body and the combined convection and radiation loses from the cylinder surface. The model is comprehensively validated to ascertaining its high predictive accuracy and the applicability under reported laser-assisted machining operating conditions. The extensive parametric analyses carried out deliver clear insight into the dynamics of thermal penetration occurring within the workpiece due to laser spot preheating. This facilitates appropriate consideration of laser preheating intensity in relation to other operating variables to achieve necessary material softening depth at the workpiece surface prior to setting out on the subsequent machining process. Building upon the data generated, a practically simpler and cost-effective preheating parametric predictor is synthesised for laser-assisted machining using neural network principles incorporating the Levenberg–Marquardt algorithm. This predictive tool is trained and verified as a practical preheating guide for laser-assisted machining for a range of operating conditions.