【作者】 张翔；

【导师】 林莉； 李喜孟；

【作者基本信息】 大连理工大学 ， 材料无损检测与评价， 2011， 硕士

【摘要】 建立含孔隙复合材料的物理模型是利用超声衰减法进行孔隙率检测的前提和关键。已有的物理模型将含孔隙复合材料假设为含有大小相同、分布均匀的规则形状气孔的弹性各向同性均匀介质,但利用该模型得到的理论预测结果与实验结果偏差较大,且存在某些实验现象难以解释的问题。基于本课题组前期在碳纤维增强复合材料(Carbon Fiber Reinforced Polymer, CFRP)孔隙形貌金相统计分析方面取得的研究结果,以及依据统计学中随机过程理论初步建立的二维随机孔隙模型(Random Void Model, RVM),本论文针对孔隙率变化范围在0.03-4.62%的15组CFRP试样,对二维随机孔隙模型(2D-RVM)建立过程中引入的粗糙度因子r、自相关长度a与b、扰动标准差ε和孔隙率P等五个参数进行了优化分析。并采用时域有限差分(Finite Difference Time Domain, FDTD)法正演模拟了超声波在RVM中的传播过程,研究了孔隙形貌对超声衰减系数和超声纵波声速的影响。对二维随机孔隙模型的优化结果表明,粗糙度因子r影响孔隙边缘光滑度,自相关长度a、b决定模型中的平均孔隙尺寸,扰动标准差ε决定孔隙与基体密度的偏离程度。参数优化后RVM中孔隙形貌统计数据与试样金相统计数据误差小于2%,优化后的2D-RVM满足数值计算精度。在保持孔隙率不变的条件下,模拟计算了孔隙形状、取向及数量等因素单一变化及孔隙形貌特征多因素随机变化下的超声衰减系数。对于椭圆形孔隙,随着孔隙在垂直于超声波传播方向上的尺寸增加,超声衰减系数不断增加。对于均匀分布的球形孔隙,孔隙尺寸越大分布越集中,超声衰减系数越大。针对0.6-4.62%之间8个不同孔隙率,对孔隙形貌与孔隙率、超声波频率和超声散射衰减系数之间的关系研究表明,形貌的随机变化会引起超声衰减系数波动,且孔隙率越大,衰减系数波动越大。此外,异常大尺寸的孔隙对散射衰减系数有较大影响。针对0.03-4.62%的15个孔隙率试样统计结果,每个孔隙率分别采用9组不同自相关长度a、b建立2D-RVM并模拟计算超声纵波声速,讨论分析了孔隙率P、自相关长度a和b、自相关长度比b/a与超声纵波声速的关系。计算结果表明,孔隙率越大,超声纵波声速越小；随着自相关长度a或b增加,纵波声速分别呈现小幅下降或上升趋势,孔隙形貌变化对超声纵波声速的影响十分微弱。更多还原

【Abstract】 It’s of great significance to establish the physics model with void compound materials in order to carry out the void content test with the method of ultrasonic attenuation. In the present physics model void compound materials were supposed to be formal medium consisting of blowholes with the same size and regular shape. But the prediction concluded from the model greatly differed from the experimental results and there was some results not easily to be explained. Based on the research results on the metallurgical statics of void morphology in the case of Carbon Fiber Reinforced Polymer and the established 2-dimensional Random Void Model resulted from the random theory of statics, in this study five parameters introduced in the establishing process of 2-dimensional RVM were analyzed and optimized:roughness factor r, autocorrelation length a and b, disturbance standard deviation s and void content P which was aimed at 15 sets of samples with porosities ranging from 0.03% to 4.62%.In addition the ultrasonic transmission process in RVM was also simulated with the method of Finite Difference Time Domain and the effect of void morphology on the ultrasonic attenuation coefficient and the ultrasonic longitudinal wave velocity was also investigated.Optimization on the 2-dimensional Random Void Model showed that void margin smoothness, the average void size of the model and the deviation between the void and the matrix were all determined by roughness factor r, autocorrelation length a and b and disturbance standard deviationε, respectively. The error between the void morphology statics of RVM and metallurgical statics of samples was less than 2% after optimizing parameters, which met the value calculation precision.For the unchanged void content, the ultrasonic attenuation coefficient affected by single morphological factor such as the shape, orientation, amount and distribution as well as multiple random characters of voids was calculated in the simulation. For the void with ellipse shape, the ultrasonic attenuation coefficient both increased with increasing size of void vertical to the ultrasonic transmission direction. Aimed at the 8 void contents ranging from 0.6 to 4.62%, it was found that the random change in void morphology would result in the fluctuation of ultrasonic attenuation coefficients and higher void content level led to larger fluctuation of ultrasonic attenuation coefficients and unusual large voids would take significant effect on the scattering attenuation coefficients after investigating the void morphology and its correlation with porosity volume fraction, ultrasonic frequency and ultrasonic scattering attenuation coefficient.Aimed at the statistics results for 15 void contents ranging from 0.03 to 4.62%,2D-RVM was established by using nine pairs of autocorrelation length a and b for each porosity and the ultrasonic longitudinal-wave velocity was simulated and calculated. The relationship between the ultrasonic longitudinal-wave velocity and the void content P, autocorrelation length a and b and autocorrelation length ratio b/a was also discussed and analyzed. Calculated results showed that the larger the void content was, the smaller the ultrasonic longitudinal-wave velocity was. With autocorrelation length a and b tending to go up, correspondingly, the acoustic velocity would slightly decrease or increase. The change of void morphology would have an effect on the longitudinal-wave velocity.更多还原