SEDL / STP / STP626-EB / STP27086S

Stress-Rupture Data Correlation with Generalized Regression Analysis: An Alternative to Parametric Methods

Rummler, DR
Materials engineer, Langley Research Center, National Aeronautics and Space Administration, Hampton, Va.

Pages: 17    Published: Jan 1977

Download this paper for $25 PDF (236K)          View License Agreement
Abstract

The applicability of multiple regression analysis techniques to stress-rupture data correlation has been investigated. A generalized interacting variable (GIVAR) method of data correlation is proposed and evaluated. The GIVAR method is compared to 6 parametric methods of data correlation on 20 sets of data. In all cases, the GIVAR method provided the best data correlation. Application of prediction intervals is discussed. The applicability of correlating variables, such as tensile strength and elastic modulus, in addition to temperature and stress, is also discussed.