节点文献

外圆纵向磨削表面粗糙度的在线测量研究

Research of On-line Measurement of Surface Roughness during Cylindrical Grinding Process

【作者】 孟繁禹

【导师】 王龙山;

【作者基本信息】 吉林大学 , 机械制造及其自动化, 2004, 硕士

【摘要】 随着科技的进步,人们对零件的制造精度和表面粗糙度的要求不断提高。外圆磨削加工往往是轴类机械产品的终级加工工序,它的应用范围越来越广,已在高效及高精密加工领域中占有重要的地位。长期的生产实践表明,降低零件的表面粗糙度,对保证配合的可靠性和稳定性;对减少摩擦系数,降低动力消耗;对提高机械和仪器的工作精度和灵敏度;对增大支撑面积,减少磨损,提高接触刚度;对减少应力集中,增加耐疲劳强度,减少振动和噪声等均起着重要的作用:即表面粗糙度直接影响着机械和仪器的使用性能和使用寿命。因此,在保证零件尺寸、形状和位置精度的同时,决不能忽视表面粗糙度。表面粗糙度的测量可分为接触式和非接触式测量两种方式。接触式测量以触针式仪器为代表。它是用一种特殊触针以一定的速度沿着被测工件表面移动,由于表面的微观不平引起了触针的上下运动,并把触针移动的变量通过机械的、光学的、电学的转换,再经放大、运算,由指示表指示被测表面粗糙度的评定参数数值,或用记录仪描绘微观不平轮廓的一种检测方法。然而,针描法轮廓仪的缺点是:由于它受到触针尖半径大小和测量速度的限制,不能用于测量R<0.012的表面粗糙度(即13级以上的表面粗糙度);此外,因金刚石触针性脆,故不适于测量R>5.0的表面粗糙度(即5级以下的表面粗糙度)。用针描法对一个被测表面进行测量,不可能完全避免划伤,因此,针描法不适合对软质材料、精加工后的工件进行粗糙度的测量。可以考虑采用非接触式测量方式。在表面粗糙度非接触测量中 ,光学法是测量的主体,其测量精度高,适于对软质材料,易损工件,以及某些带有有用信息的表面进行测量。采用光学原理的表面粗糙度非接触测量方法(如光切法﹑光波干涉法﹑散斑法﹑反射光能法即反射散射比法等)则可以较好弥补触针式仪器的不足。 <WP=74>为了提高表面粗糙度的检测效率,可以采用光学法进行在线测量方式。在线测量是一种实时的快速自动测量,是指在加工过程中的其它时间,如在流水线生产中当转移工位时(加工零件仍旧装夹在机床上),定时地无间断地进行的测量。表面粗糙度的在线测量,不但能正确的检测加工过程中参数状态的变化,而且对控制和改进加工方法、提高加工表面的质量和产品性能,有着十分重要的意义。本文利用激光图谱比较法进行表面粗糙度的在线测量。物体被高度相干的激光照射后,具有一种独特的颗粒状的外貌,这种颗粒状态称为激光散斑,亦称激光光斑。光学散斑法(又称光斑法)是利用平均散斑对比度测量表面粗糙度有关参数的一种方法。当以一定角度入射的激光束照射到粗糙表面上时,将会按照光学定律产生散射和反射现象。表面粗糙度高度参数越小的表面,镜面反射的中心亮斑越强,散射的光带越弱,并且光带随之变短变窄;反之,表面越粗糙,中心亮斑越弱,散射光带越强。在入射光的反射方向上,使用面阵CCD摄像机进行接收。被摄物体反射的光线,传播到镜头,经镜头聚焦到CCD芯片上,CCD根据光的强弱积聚相应的电荷,经周期性放电,产生表示一幅幅画面的电信号,经过滤波、放大处理,通过摄像头的输出端子输出一个标准的复合视频信号。该视频信号经计算机进行图像处理后,输出表面粗糙度等级。通过对外圆磨削表面粗糙度标准样块进行多次测量,以此统计数据建立数据库,再对外圆纵向磨削加工后的零件进行表面粗糙度非接触式在线测量,将此测量结果与数据库中的数据进行比较,从而判定加工零件表面粗糙度的等级。表面粗糙度检测,可以考虑通过对图像纹理分析的方法来实现。纹理表现为所观察到的图像子区域的灰度变化规律。一些物体表面可能具有与方向相关的纹理信息。一般来说,纹理图像中灰度分布具有某种周期性,即使灰度变化是随机的,它也具有一定的统计特征。我们可以直观地认为图像纹理结构具有以下三个特征:①是某种局部的序列性在该序列更大的区域内不断重复;②是该序列由非随机排列的基本部分组成;③是纹理区域内大都是均匀的统一体,图像往往显示出重复性结构。本文利用傅立叶变换,可以把空间域中需要解决的问题转换到空间频率域中去解决:由于被检测件表面成像是一幅二维图像,因此采用二维离<WP=75>散傅立叶变换方法使之转换到频域中对信号进行谱分析。通过对能量谱的分析,区分判别不同等级的表面粗糙度。如果一幅图像的纹理较粗糙,即图像的灰度变化很少或较慢,则在小的值处| F(u, v)|应有较大的值;如果一幅图像的纹理较细腻,即图像的灰度变化频繁或较快,则在大的值处| F(u, v)|应有较大的值。因此,若检测纹理的粗糙、细腻性质,可以利用建立| F(u, v)|(能量谱E)随(能量谱所在半径R)变化的关系进行判断。本文把所有在线测得工件的图谱数据交给神经网络,通过使用图像识别技术,与数据库中存储的图像进行对比识别,并输出被测件的表面粗糙度等级。这种算法能排除人为因素,正确识别不同的表面,并具有较高的精度。本文使用的测量方法,可以在比普通磨削效率提高3~4倍的前提下,实现在线检测表面粗糙度的目的。采用激光图谱比较法检测工件测得的结果,通过与触针式轮廓测量仪进行

【Abstract】 With the development of science and technology, the requirement for fine tolerances and roughness of components is constantly improved. Cylindrical grinding as a final procedure in the productions of precision metal process , used by many fields, it has occupied an important place in efficient and high tolerances field.Decreasing the parts’ surface roughness, as it is shown after a long time practice that can realize the reliability and stabilization of the match; can reduce faction coefficient, wear, stress concentration, vibration, noises, and power consumption; can improve the operative precision and perceptivity; can increase the support area and contact intensity; and also can improve wear degree etc. The surface roughness affects capability and longevity of machine and apparatus directly. When we ensure the parts’ dimension, shape and position precision, we should not neglect the parts’ surface roughness.The measurement of surface roughness has two methods: the contact and the non-contact. The contact measurement uses a special needle moving along the parts’ surface with a certain speed. The minute wave of the surface causes the needle to move up and down. After mechanical, optical and electrical conversion, and then amplified, operation, the variation of the needle can show the value of the surface roughness parameter by some instruments or the minute wave profile can be described by some record instruments. However, the contact measurement has many disadvantages: because of the restraint of needle’s radius and the measurement speed, it can not be used for measuring the surface roughness value R<0.012. Furthermore, because the needle is <WP=77>crumbly, it is not fit for measuring the surface roughness value R>5.0. Finally, it is impossible for the contact measurement to measure a surface without any derogation. In short, the contact measurement is not fit for measuring soft material or parts’ surface roughness processed by a precision metal process.In the non-contact measurement method of the surface roughness, optical measures are major. Because they have high precision and are fit for measuring soft materials, the easily scathed parts and some surfaces with some useful information. The non-contact measurement can compensate the contact measurement’s disadvantages very well.In order to increase measurement efficiency of the parts surface roughness during cylindrical lengthways grinding process, an on-line optical measurement method can be adopted. On-line measurement is a kind of on-time measurements. And it can realize constantly measurement during the else time of the processing courses.The on-line measurement of the surface roughness can not only monitor the changes of the state parameters during the processing courses correctly but also can be important for controlling and improving process methods, the qualities of processing surface and the capabilities of the products.Exploiting laser spectrum comparing method, a surface roughness measurement is carried out. When a beam of laser is poured on a rough surface, it will produce a kind of specific particulate appearance. This is called scattered laser spots. According to the optical laws, it will produce scattered and reflection phenomena. The surface’s roughness altitude parameter is smaller, the center illumination spots of reflection are more intensive, the scattered radiation belt becomes weaker, shorter and narrower. The surface is rougher, the center illumination spots of reflection are weaker and the scattered radiation belt becomes stronger. In the direction of the reflection, we use a CCD camera to receive the beam. The received reflection beam is focused on a CCD chip by a lens. According to the intensity, the CCD concentrates corresponding charges. It can produce a map of electrical signals after it releases charges periodically. The signals can be changed into standard <WP=78>multiple vision signals after they are filtered and amplified. The vision signals can output the degree of roughness after they are approached

  • 【网络出版投稿人】 吉林大学
  • 【网络出版年期】2004年 04期
  • 【分类号】TG84
  • 【被引频次】6
  • 【下载频次】730
节点文献中: 

本文链接的文献网络图示:

本文的引文网络