Abstract: Aiming at investigating the limitations of weak cooling effects to difficult cutting materials under minimum quantity lubrication (MQL) conditions, the influences of cooling parameters on tool vibration and surface roughness under MQCL conditions were analyzed. The orthogonal test was designed based on Taguchi method, and cutting test was carried out based on MQCL conditions, then the effects of cooling parameters (cold air temperature, oil flow rate, wind speed and jet surface type) on tool vibration and surface roughness were analyzed by the analysis of variance, main effects plot, response surface methodology and metal-cutting principle. The prediction model of tool vibration and surface roughness were established related to parameters. The improved genetic algorithm was used to optimize the support vector regression (SVR)prediction model synchronously to obtain the optimal values of cooling parameters. The experimental results show that the temperature has the greatest influence on the tool vibration, and the tool vibration increases with the increase of temperature. The influences of wind speed on the surface roughness are the greatest, when the wind speed is less than 10 m/s, the surface roughness increases with the increase of wind speed, when the wind speed is greater than 10 m/s, the surface roughness decreases with the increase of wind speed. When the surface to be jetted is the toos minor flank, the surface roughness and the vibration are the least. The cooling parameter optimization results show that， when the cold air temperature is as -2.36 ℃, the wind speed is as 7.31 m/s, the oil flow is as 300 mL/h, and the spraying surface is the tool minor flank, the surface quality of the workpiece is the best, and the value of surface roughness is as 0.6588 μm. The verification experimental results show that the prediction errors of surface roughness and RMS of vibration are as 4.4% and 5.9% respectively.

Key words: minimum quantity cooling lubrication (MQCL), surface roughness, tool vibration, genetic algorithm-support vector regression（GA-SVR） optimization

摘要： 针对最小量润滑(MQL)技术在加工难切削材料时冷却能力不足的问题，分析了最小量冷却润滑(MQCL)条件下冷却参数对刀具振动和表面粗糙度的影响规律。设计了以田口法为基础的正交试验方案，并基于MQCL条件进行了相关切削试验。采用方差分析法、主效应图法、响应面法等方法并结合切削理论，分析了冷风温度、油液流量、风速、喷射面类型等冷却参数对刀具振动和表面粗糙度的影响机制，建立了与冷却参数关联的加工刀具振动和表面粗糙度预测模型，同时采用改进的遗传算法对支持向量回归预测模型进行同步优化，得到冷却参数的最优值。试验结果表明，温度对刀具振动的影响最大且随着温度升高刀具振动呈现出增大的趋势；风速对表面粗糙度的影响最大，当风速小于10 m/s时，随着风速增大表面粗糙度增大，当风速大于10 m/s时，随着风速增大表面粗糙度减小。当喷射面为刀具副后刀面时，刀具振动和表面粗糙度均最小。冷却参数优化结果表明，当冷风温度为-2.36 ℃、风速为7.31 m/s、油液流量为300 mL/h、喷射面为副后刀面时，工件表面质量最好，其表面粗糙度Ra为0.6588 μm。验证实验表明，表面粗糙度和振动均方根的预测误差分别为4.4%和5.9%。

关键词: 最小量冷却润滑, 表面粗糙度, 刀具振动, 遗传算法支持向量回归优化