| At present,die materials regardless of its variety,specification,quality and quantity may basically satisfy the needs of modern mold manufacturing.But for a hot-working die which has most complex load and worst working conditions,educing the material which has best using performance,manufacturability and heat treatment is always the key point that the researchers care about.In the traditional material and heat treatment selection process,the experts’ experiences are major relied.This method is bound to make great limitation,especially in nowadays that a great number of new materials emerged.So it is necessary to update this traditional and log method.The research purpose and significance of this thesis is how to combine die’s material and heat treatment selection with decision support technology and artificial intelligence and how to develop the intelligence decision-making support system which is adapt to the field of die design and manufacture and possess independent intellectual property for die material and heat treatment selection.Making full use of the achievement of primary researches and financed by the International Cooperation Item of Jiangsu Province and the Jiangsu University Die Technology Innovation Grope Item,the hot-working die material selection and heat treatment intelligence decision-making support system which possess independent intellectual property is firstly developed based on summarization of the existing hot-working dies’ working conditions and heat treatment process and combined with the advantages of expert system and decision support system.The main research content and innovative points in this paper:1.According to hot-working die failure resistance system,the hot-working die material selection intelligence decision-making model."Expectation Attribute Comprehensive Value"(Qx) is innovatively established to evaluate material’s comprehensive service performance in this model.Therefore,existing evaluation system in hot-working die material and heat treatment selection which is rough and partly quantized is successfully changed.2.Based on a great of documenting work,information entropy has been used to compute the weight of material properties and TOPSIS has been used to evaluate the Qx value of the pre-selected materials.Then the best material and heat treatment selection is obtained.3.We take thermal fatigue as research objects which is the major failure mode of hot-working die.The relative material properties and thermal fatigue performance are respectively used as input and output variables.Partial least square(PLS) is used to build thermal fatigue prediction model which can predict material’s thermal fatigue performance rapidly and accurately.It has important theoretical guidance meaning and engineering value to new hot-working steel development. |