纵-弯复合振动超声挤压加工试验研究

为研究挤压加工时静压力、进给量和挤压速度等工艺参数对试件表面质量的影响,在挤压加工中引入纵-弯复合振动后对Q235钢轴件端面进行处理,并与普通挤压加工进行对比,基于正交试验结果构建了挤压加工后表面粗糙度和表面里氏硬度二次回归预测模型.试验发现:在相同加工工艺参数下,在普通挤压加工中引入纵-弯复合超声振动后获得的表面粗糙度Ra值更小,而表面里氏硬度值显著提高;采用两种挤压加工方法后工件表面粗糙度Ra值均随着静压力和进给量的增加而增大,而挤压速度的影响很小,进给量对表面粗糙度的影响最为显著;工件经超声挤压加工时静压力越大,则获得的表面硬度越大,且表面硬度随进给量的增大先增大后减小,而普通挤压加工后表面硬度随静压力和进给量的增大先增后减,且在两种加工方式下工件表面硬度基本不受挤压速度的影响.纵-弯复合振动超声挤压加工工艺适合Q235钢表面强化处理,构建的表面粗糙度和硬度的预测模型可用于指导Q235钢表面强化处理工艺生产.

Experimental study on ultrasonic extrusion with longitudinal-flexural vibration

Longitudinal-flexural vibration was intraduced into conventional extrusion (CE) to deal with the Q235 steel in order to study the effects of the processing parameters on the surface roughness and the hardness. And the prediction models of the surface roughness and the hardness for ultrasonic extrusion (UE)Q235 steel with longitudinal-flexural vibration were constructed in the light of the onthogonal experimental results using the quadratic regression analysis method. The results show that compared to CE, the surface roughness R_a obtained by UE are smaller, while the hardness HL are much higher than that of CE with the same processing parameters. In the process of UE and CE, the value of R_a increases with the static pressure and the feed-rate increase, and changes little with the increment of the extrusion speed. And the feed-rate has greatly impact on the surface roughness. The hardness HL obtained by UE increases with the increment of the static pressure, and when the feed-rate increase, the hardness first increase and then decrease. However, the hardness obtained by CE increases firstly and then decreases with the increments of the static pressure and the feed-rate. Meanwhile, the extrusion speed has little effect on the hardness in the process of UE and CE. The prediction models of the surface roughness and the hardness are significant at the 90% confidence level, and the optimal processing parameters obtained by the prediction models are similar to that obtained by the orthogonal test, indicating that the prediction models are reliable.