An efficient load identification for viscoplastic materials by an inverse meshfree analysis

• Published in: International journal of mechanical sciences Vol. 136; pp. 303 - 312
• Main Authors: Kazemi, Z., Hematiyan, M.R., Shiah, Y.C.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.02.2018
• Subjects: Cartesian transformation method (CTM); Damped Gauss–Newton method; Inverse analysis; Load identification; Meshfree radial point interpolation method (RPIM); Viscoplastic; Viscoplastic; Load identification; Inverse analysis; Meshfree radial point interpolation method (RPIM); Damped Gauss–Newton method; Cartesian transformation method (CTM)
• ISSN: 0020-7403; 1879-2162
• DOI: 10.1016/j.ijmecsci.2017.12.050

•An inverse method for load identification in viscoplastic problems is presented.•The unknown load has variation with respect to both space and time.•Strains at some sampling points are used as measured data in the inverse analysis.•Sampling points can be located on boundary or within the domain of the problem.•The method gives satisfactory results for a high level of noise in measured data. Despite the extensive study of direct viscoplastic analysis in the past, its inverse study has remained very scarce indeed. In this paper, an inverse method based on an improved version of the meshfree radial point interpolation method (RPIM) is presented for load identification in 2D viscoplasticity. The unknown load, varying with respect to space and time, is determined using measured strains at several sampling points on boundary or within the domain of the problem. The inverse analysis employs the well-known Tikhonov regularization and damped Gauss–Newton methods. Proper location and arrangement of sampling points for more accurate identification of unknowns is investigated too. To demonstrate the feasibility of the proposed method, a comprehensive numerical example in different conditions is presented. Furthermore, the effects of some important parameters, such as the number of sampling points and measurement errors, on the stability and accuracy of the solution are also studied. Despite the extensive study of direct viscoplastic analysis in the past, its inverse study has remained very scarce indeed. In this paper, an inverse method based on an improved version of the meshfree radial point interpolation method (RPIM) is presented for load identification in 2D viscoplasticity. The unknown load, varying with respect to space and time, is determined using measured strains at several sampling points on boundary or within the domain of the problem. The inverse analysis employs the well-known Tikhonov regularization and damped Gauss–Newton methods. For more accurate identification of unknowns, proper location of the sampling points is investigated too. To demonstrate the feasibility of the proposed method, a comprehensive numerical example in different conditions is presented. Furthermore, the effects of some important parameters, such as the number of sampling points and measurement errors, on the stability and accuracy of the solution are also studied. [Display omitted]