中文摘要：

对我国6个试验点17种钢的16年大气腐蚀试验数据进行幂函数回归，将所得到的幂函数参数A、n对试验钢的化学成分和试验的环境数据进行逐步回归，得到钢的大气腐蚀发展幂函数的参数与钢的化学成分和环境因素的定量关系．由此根据钢的化学成分和所在环境的主要气象和污染数据。就可以预测碳钢、低合金钢在任何给定环境下的大气腐蚀状况，模型显示：降雨量和日照时数的乘积是一个特殊的重要影响因素，它降低A值，提高n值．钢中的S元素是最有害的，它对A、n值都有较高的正系数；P对大气腐蚀是有益的，对A、n值都有负系数；Cu只降低了n值，而对A值几乎没有影响；Mn和Si的影响与Cu相反，它们只降低A值，而不降低n值，Si甚至提高n值．

英文摘要：

Multi- argument linear stepwise regression were made for the atmospheric corrosion data of steels after sixteen- year exposure at various sites in China to analyze the effects of various factors on the corrosion. First, power function was taken in the regression for the weight loss data. The obtained parameters of the power function were then used as the variants in the stepwise regression. Both chemical compositions of steels and the environment factors were used as the arguments in the stepwise regression. Quantitative relation of the corrosion parameters with both environment factors and chemical compositions of steels was obtained. Apart from sulfur dioxide, marine salt and humidity, the well known three damaging pollution factors, the rainfall and sunshine hours together is a special environment factor, and it has a positive effect on n value but a negative effect on A value. A high rainfall together with a high sunshine hour resulted in a low corrosion rate at the initial stage and a high corrosion rate in later years. About the effect of alloy elements on the corrosion resistance of steels, S is the most damaging element in steel for atmospheric corrosion. P is a full beneficial element. Copper reduces only the n value but has little effect on the A value. The effects of Mn and Si are opposite to that of Cu. They reduce only A value but not n value. Si even increases the n value. Corrosion can be predicted for most carbon and low alloy steels in various environments.