钢材低温力学性能变化规律的灰色预测方法

The Grey Prediction Method for Mechanical Behaviors of Steel at Low Temperature

目的验证灰色预测系统非等间距GM(1,1)模型对钢材低温力学性能预测的准确性与有效性,以保证钢结构工程服役于低温工作环境的安全性与可靠性.方法应用灰色相关理论,提出基于积分构造背景值的非等间距GM(1,1)预测模型,并用MATLAB对所构建的预测模型编写程序,以常用的钢材(Q235、16Mn、15Mn V、16Mnq、14Mn Nbq)在低温下的力学性能试验数据为初始序列进行灰色预测试验研究.结果力学试验结果表明钢材在低温工作环境中,其主要力学性能指标:屈服强度fy和极限强度fu随温度的降低而提高;断后伸长率δ和断面收缩率ψ随温度的降低而降低;预测结果显示该非等间距GM(1,1)模型预测精度均为一级,小概率误差均为1,最大相对误差2.5%,数据拟合精度高达97.5%.结论非等间距GM(1,1)模型对钢材低温力学性能预测方法精度高,具有很好的工程应用价值.

To ensure safety and reliability of steel structures under low temperature,the accuracy and efficacy of a proposed prediction unequal GM（ 1,1） interval model of grey prediction system w ere verified in this paper. Applying the grey theory,based on the integral structure background value,a prediction unequal GM（ 1,1） interval model w as proposed and a M ATLAB program for this model w as built. M echanical behavior test data of frequently used steels（ Q235,16 M n,15 M n V,16 M nq,14 M n Nbq） under low temperature w ere taken as a initial sequence to conductgray predict investigation. M echanical results show that w hen w orking under low temperature,the yield strength fyand ultimate strength fuincrease w ith decrease of temperature; w hile the percentage of elongation δ and area reduction ψ after fracture decrease w ith decrease of temperature. The error of small probability is alw ays 1; the maximum relative error is 2. 5%; the data fitting precision of the prediction mode is quite high up to 97. 5%. Conclusion is that the proposed unequal GM（ 1,1）interval model is of certain accuracy and efficacy for mechanical behavior prediction of steel under low temperature. It can be seen that this model is valuable for practical engineering application.