结构冲击畸变问题的直接相似方法研究

The distortion of structural impact arising from the use of different materials between the full-size protype and the small-size scaled model is usually compensated by the correction for the velocity or density of the scaled model. However, the traditional correction methods are limited for the application of similarity laws because of some inherent defects such as the needs to test structural response in advance, the dependence for the special constitutive equation and the inability to reflect dynamic process of impact. In the present paper, a direct scaling method for distortion problems is proposed. Based on the best approximation relationships between the predicted flow yield stress of the scaled model and the flow yield stress of the prototype on strain rate interval, the scaling factor for the corrected velocity or the corrected density are obtained directly, and the dynamic similarity relations between the scaled model and the prototype are established. Based on the Norton-Hoff, Cowper-Symonds and Johnson-Cook constitutive models, the influences of material parameters of the strain-rate-sensitive, the reference strain-rate, the yield stress and the density for the dynamic similarity relations are studied. And a crooked plate under impact mass is used to verify the effectiveness of the proposed directly scaling method. The analysis shows that the proposed directly scaling method does not need to test the structural response information in advance, does not depend on the specific form of constitutive equation and emphasis the dynamic similarity characteristics, so that it has the direct, efficient and universal characteristics. The optimal approximation effects of the dynamic similarity relations are mainly controlled by the material strain-rate-sensitive characteristic parameters, while the reference strain-rate, the yield stress and the density have little influence. When the material strain-rate-sensitive characteristic parameters of the scale model are similar to those of the prototype, the best dynamic approximation effects can be obtained.