Statistical Analysis On Creep Rupture Property Of Heat-Resistant Steel

High temperature components have been widely used in petrochemical industry,power plant,aerospace industry and other applications.After long-term service in high temperatures and pressures simultaneously,their creep rupture properties are greatly decreased due to microstructure degradation and the accumulation of damage.Because of the need in safety issue as well as in economizing fund by maintaining the existing construction instead of buying new one,creep rupture property assessment and life prediction technology of heat-resistant steels are important topics to study at home and abroad.By reasonably considering the dispersibility of material creep rupture data,Z parameter method has achieved life prediction with reliability for heat-resistant steels.This paper makes statistical analysis for creep rupture data of tens of heat-resistant steels using Z parameter method.The results show that the scattering of Z parameter for these heat-resistant steels is supported by normal distribution.And reliability analysis method based on Z parameter can be used in creep rupture data analysis and assessment for these steels.Stress-TTP-Reliability curves(σ-TTP-R curves),Stress-Rupture Time-Reliability curves(σ-t_r-R curves) and Allowable Stress-Temperature-Reliability curves([σ]-T-R curves) are proposed for heat-resistant steels based on Z parameter in this paper,which could embrace reliability concept into creep rupture property design.Because the constant selection of time-temperature parameter has obvious influence on the result of creep rupture life prediction,this paper analyze the selection problem for Larson-Miller parameter and Manson-Haferd parameter,which are widely used in engineering.For Larson-Miller parameter model,there are obvious changes in the distribution trend of creep rupture data and life prediction results when different constant C value is selected.However,for Manson-Haferd parameter model,constant T_a and lgt_a are phenomenally not independent variable but show a linear relationship:lgt_a=27.015-0.0267T_a. Desirable correlation results and prediction accuracy can be obtained even the values of Manson-Haferd constants(T_a,lgt_a) are selected in a broad range on the relation line.Unified Manson-Haferd constant(450,15) can be selected to analyze creep rupture data and get prediction results with high accuracy for many steels.This paper simultaneously compares reliability analysis method and the safety coefficient method in creep rupture data analysis and design for heat-resistant steels.The reliability analysis method based on Z parameter is also compared with minimum rupture stress method used in Petroleum and Chemical Industry Standard(SH/T 3037-2002).