Search

Abstract: A general numerical model was presented to analyze the interval inverse hyperbolic heat conduction problem, with Bregman distances and weighted Bregman distances as regularization terms. By using the interval finite element method and interval extension theory, the direct and inverse models were established for uncertainties. The eight-point isoparametric elements were applied for the discretization in the space domain, and the Precise algorithm in time domain was empolyed. The inverse problems were implicitly formulated as optimization problems, using squared residues between the calculated and measured quantities as the objective function of the inverse identification. Results show that the proposed numerical models can identify single and combined interval thermal parameters and boundary conditions for hyperbolic heat transfer problems accurately and efficiently. [Copyright &y& Elsevier]

• The modified Corten-Dolan model excellently captures the load interaction effect. • Generalized expression of the new model gives a unified theoretical formulation. • The new-introduced validation scheme comprehensively investigates applicability. • The new model has much better prediction ability than the existing models. Corten-Dolan nonlinear damage accumulation model can well describe the fatigue behavior of engineering materials and structures under variable amplitude loads, while the determination of parameter d is difficult and still open. In this paper, a new improved version of Corten-Dolan model is proposed in which the parameter d is modified by introducing the power form of the stress ratio between two consecutive load levels to reflect the nonlinear effect of load interaction. Furthermore, a generalized expression of the new model is developed, which gives a unified theoretical formulation for the presented and some typical existing versions. The proposed model is validated by comparison with existing versions based on the experimental data of typical engineering materials under two-level and multilevel loads. An effective validation scheme is introduced to comprehensively investigate the validity and applicability of the developed model. In this scheme, parts of data those are not used in modeling are regarded as the check data to examine the prediction accuracy. The results show that the developed model can well reflect load interaction effect and give more satisfactory prediction of fatigue life. [ABSTRACT FROM AUTHOR]