New framework for Bayesian statistical analysis and interpolation of residual stress measurements |
| |
Affiliation: | 1. Department of Mechanical Engineering, Solid Mechanics, Technical University of Denmark, Kgs. Lyngby DK-2800, Denmark;2. Center for Bachelor Engineering Studies, Technical University of Denmark, Ballerup DK-2750, Denmark;1. Institute of Seismology and Volcanology, Faculty of Sciences, Kyushu University, Japan;2. Disaster Prevention Research Institute, Kyoto University, Japan |
| |
Abstract: | In the present study, different residual stress and strain data measured from various techniques are analyzed using a Bayesian statistical approach and finally interpolated utilizing modified Shepard method. This research is carried out to compare the capability, simplicity and accuracy of Bayesian approach with different probability density functions. Three different probability density functions: Gaussian, Cauchy and Sivia's distribution are studied and compared here. Finally the modified Shepard method is utilized with new interpolant and weight functions, to interpolate the scattered measured data. The proposed framework is then applied to two sets of measured residual data obtained from various experimental techniques. |
| |
Keywords: | Residual stress Bayesian approach Modified Shepard method Cauchy P.D.F. Weighted least square technique |
本文献已被 ScienceDirect 等数据库收录! |
|