A generalization for Parker rotational relaxation model based on variable soft sphere collision model

1996 ◽  
Vol 8 (5) ◽  
pp. 1336-1337 ◽  
Author(s):  
Katsuhisa Koura
2013 ◽  
Vol 441 ◽  
pp. 476-479
Author(s):  
Wei Wei Zhang ◽  
Hong Xu

A robust and efficient parameter identification method of the stress relaxation model based on Altenbach-Gorash-Naumenko creep equations is discussed. The differential evolution (DE) algorithm with a modified forward-Euler scheme is used in the identification procedure. Besides its good convergence properties and suitability for parallelization, initial guesses close to the solutions are not required for the DE algorithm. The parameter determination problem of the stress relaxation model is based on a very broad range specified for each parameter. The performance of the proposed DE algorithm is compared with a step-by-step model parameter determination technology and the genetic algorithm (GA). The model parameters of 12Cr-1Mo-1W-1/4V stainless steel bolting material at 550°C have been determined, and the creep and stress relaxation behaviors have been calculated. Results indicate that the optimum solutions can be obtained more easily by DE algorithm than others.


2013 ◽  
Vol 842 ◽  
pp. 482-485
Author(s):  
Wei Wei Zhang ◽  
Hong Xu

In order to determine the parameters of a stress relaxation model based on Altenbach-Gorash-Naumenko creep equations, an efficient parameter identification scheme is discussed. The differential evolution (DE) algorithm is used in the identification procedure with a modified forward-Euler scheme. The model parameters of 1Cr-0.5Mo-0.25V stainless steel bolting material at 500°C have been determined, and the creep and stress relaxation behaviors have been calculated. Comparing with a step-by-step model parameter determination technology and the genetic algorithm (GA), it shows that the DE algorithm has better convergence property and suitability for parallelization, and no need of initial guesses close to the solution. Results indicate that the optimum solutions can be obtained more easily by DE algorithm than others.


Sign in / Sign up

Export Citation Format

Share Document