【作者】 何维军；

【导师】 赵福令；

【作者基本信息】 大连理工大学 ， 机械电子工程， 2008， 硕士

【摘要】 碳纤维复合材料(CFRP)具有轻质、高强、耐高温、抗腐蚀、热力学性能优良等特点,广泛应用于航空航天领域。由于CFRP具有非均质性和各向异性等特点,其切削加工表面容易出现纤维拔出、纤维破碎、纤维撕裂、表面凹坑等独有的表面形貌特征,采用传统的二维评定标准不能真实反映CFRP的加工表面形貌特征。三维表面形貌测量是当今研究的热点,但国际上尚未建立相应的三维评定标准。而且,目前用于表征加工表面形貌特征的粗糙度参数是与尺度相关的参数,其测量值的大小随测量仪器的分辨率,采样长度的改变而改变,其测量结果不能反映加工表面形貌的固有特征。研究发现,机械加工表面具有随机、无序、自相似和自仿射等分形特性,而且分形维数是一种与尺度无关的参数,因此,使用分形理论来研究加工表面形貌成为可能。本文从CFRP加工表面的三维形貌出发,应用分形、小波理论,以加工表面SEM灰度图像为研究对象,对CFRP加工表面形貌特征进行研究,主要工作如下:(1)将分数布朗运动引入到表面形貌的研究中,分析分数布朗运动的特性和分数布朗运动模型参数H的物理意义,利用分数布朗运动模型模拟不同H值的分形曲线和分形曲面,研究了基于SEM图像的离散分数布朗随机场模型。(2)采用差分盒计数法(DBC)计算了CFRP加工表面图像的分形维数,结果表明,分形维数D与加工表面三维形貌密切相关,加工表面越粗糙,D越大,分形维数D能综合反映加工表面形貌的整体变化特性;引入多重分形对加工表面局部形貌特征进行分析,并利用计算的多重分形谱参数分析加工表面局部形貌的变化特性。(3)基于离散分数布朗随机场模型对CFRP加工表面图像纹理特征进行了研究,将图像中各点像素的灰度值变为相应的分形参数值H和b,对H进行归类,对同一类的H设定为相同的灰度值,利用设定的不同灰度值对表面图像进行纹理分割,从而提取出加工表面的局部形貌特征。(4)基于小波变换对CFRP加工表面粗糙度纹理特征的变化规律进行了分析,加工表面的粗糙度纹理变化主要反映在表面图像的高频信息部分,对表面灰度图像进行小波分解,计算分解的各细节子图像所占的能量比例,以此作为加工表面粗糙度纹理在不同方向上变化的依据。更多还原

【Abstract】 Carbon fiber reinforced plastic composite (CFRP) has many excellent properties, such as light weight, high specific strength, high temperature resistance, corrosion resistance and so on, which is widely used in aeronautics & astronautics. Because of the inhomogeneous structure and anisotropy performance for CFRP, it appeared easily with the unique surface topography characteristics such as fiber pull-out, fiber crushing, fiber tearing, et al. The traditional 2-D evaluation standard is not suitable for CFRP anymore. At present, 3-D topography roughness is the research focus, but the corresponding 3-D evaluation standard has’t been established internationally until now, and roughness parameters are relevant to scales, the measured values change with the change of the instruments’s resolution and the sampling length, so that it can’t reflect the intrinsic characteristic of the surface topogryphy.Research indicates that machined surfaces have fractal characteristics of randomness, irregularity, self-similiarity and self-affinity, and fractal dimension is a kind of parameter irrelevant to scale, so it is possible to apply fractal theory for surface topography. The author of this paper goes at the 3-D surface topography of CFRP, applies the fractal & wavelet theory, takes the SEM image of the cutting surface as the research object, studies the surface topography characteristic of CFRP, the main studies are as follows:(1) The fractal Brownian motion (FBM) is introduced to the research of surface topography, the characteristic of FBM and the physical meaning of fractal parameter H are analyzed, and fractal curves and fractal surfaces are simulated with different H. And the Discrete Fractal Brownian Random Field (DFBRF) model which based on SEM image is studied.(2) The surface fractal dimensions are calculated with the differential box-counting (DBC) method. The results show that the fractal dimension D are closely related 3-D topography of the machined surfaces of CFRP, the rougher the surfaces, the bigger the D, fractal dimension can synthetically evaluated the whole surface topography characteristic. In order to analyze the local topogryphy characteristic of CFRP cutting surfaces, multi-fractal is introduced, the multi-fractal spectrums of the images are calculated, and the multi-fractal spectrum parameters are used to analyze the partial topography change characteristic of the cutting surfaces. (3) The study on the surface texture feature of CFRP is based on DFBRF, the gray values of the SEM images are changed into the fractal parameters H & b, and then fractal parameters H are classified, the same gray value is endowed with when H is in one class, the image surface texture is segmented into different blocks by the new different gray value, so that the local topography characteristics of CFRP cutting surfaces are extracted.(4) The study on the change of the surface roughness texture for CFRP is based on wavelet transform. The SEM image is decomposed by using Daubechies wavelet, the energy proportions of the decomposed subimages are calculated, and the energy proportions of the decomposed high frequency subimages are regarded as the change basis of the surface roughness texture in defferent directions.更多还原