Engineering Science and Technology, an International Journal (Apr 2024)

Instantaneous contact area-based model for shear strength sensitive cutting coefficients characterization of anisotropic parts

  • José David Pérez-Ruiz,
  • Luis Norberto López de Lacalle,
  • Gorka Urbikain,
  • Hugo Álvarez,
  • Jovanny Pacheco

Journal volume & issue
Vol. 52
p. 101650

Abstract

Read online

The anisotropic nature of laser powder bed fusion (LPBF) parts means that the cutting force coefficients obtained by linear regression or non-linear optimization methods only work for one tool position relative to the workpiece. These methods must obtain a new set of coefficients for each tool position relative to the workpiece, which becomes cumbersome for surface machining with multiple tool-to-workpiece orientations, which is typical for LPBF parts. As a first step to overcome this drawback, this paper presents a new methodology for calculating cutting coefficients sensitive to material shear strength changes when the tool changes orientation. To this end, a vector inverse model based on the instantaneous contact area is presented. The shear coefficients obtained from the present model make physical sense and provide the basis for predicting cutting coefficients for multiple tool orientations from recently developed microstructure-based models. For the model verification, prismatic IN718 LPBF samples were manufactured and machined with a peripheral endmill for three tool orientations (cases). The material was characterized with SEM and EBSD, and the plastic anisotropy was verified. The coefficients obtained with the proposed model presented a high correlation (0.97) with the cutting resistance in contrast to those obtained with the linear inverse model, which establishes that the vector-based mechanistic coefficients are more appropriate for anisotropic materials. The simulated cutting forces from the coefficients obtained through the proposed model were compared with the experimental forces, observing similar patterns and levels. The coefficients obtained with the model allow average forces to be obtained with a prediction error between 5 and 15 %, depending on the percentage of instantaneous data used to estimate the mean value of the coefficients.

Keywords