摘要
对LC9-CS棒材进行轴向应变控制疲劳试验,较全面地测试出材料的应变疲劳特性,包括稳态应力应变曲线,瞬态应力应变曲线,Manson-Coffin公式描述的εa-Nf曲线以及当量应变寿命(εeq-Nf)曲线。并对上述特性曲线构成的不同寿命预测方法与试验进行了深入比较,结果表明,瞬态应力应变曲线和εeq-Nf曲线能较好地反映材料的疲劳特性。并能兼顾应变疲劳和应力疲劳分析的需要。
LC9 aluminum alloy is but recently used in Chinese aeronautical industry. Our fatigue property tests of LC9 may very likely be the most extensive in China, numbering 192 specimens. Besides, a novel neural network method is used in analyzing test data.in this paper, a series of axial strain-controlled fatigue tests of LC9-CS bar specimenshas been conducted to get the fstigue characteristics of this material. The tests lead to a stable cyclic σ~ε curve, a number of cyclic transient σ~ε curves, a εa~Nf relationship (deduced from Manson-Coffin equation) and a εeq~Nf relationship. Especially, a set of Back Propagation Neural Network (BPNN ) is established to reflect the highly nonlinear feature of transient cyclic σ~ε curves, which is very hard to be descrigbed with traditional mathematical methods. The computation results of BPNN are satisfactory from the viewpoint of engineering application. Combining transient cyclic σ~ε curyes with εeq~Nf relationship is the best wad to describe the fatigue characteristics of material.
出处
《西北工业大学学报》
EI
CAS
CSCD
北大核心
1996年第3期376-380,共5页
Journal of Northwestern Polytechnical University
基金
国家自然科学基金
航空基金
关键词
应变疲劳
应力应变曲线
应变寿命曲线
铝合金
strain-controlled test, cyclic σ~εcurve, εeq~Nf relationship, back-propaga tion neural network