Research On The Quantitative Characterization Method Of The Microstructure Of Short Fiber Reinforced Composite Based On Microscopic Image

The microstructure of short fiber reinforced composites can be reflected by the distribution of fiber length and fiber orientation,which play an important role in analyzing the macroscopic properties of the composites.However,the traditional measurement methods of fiber length and fiber orientation based on microscopic image processing have many shortcomings,such as hefty workload,low efficiency and reliability.Therefore,the study of more convenient and reliable methods to obtain the fiber length and fiber orientation distribution of short fiber reinforced composites has become a very essential subject.An efficient and reliable method for the automatic measurement of fiber length based on skeleton segments grouping is proposed,which mainly solves the problem of fiber separation in the case of complex fiber intersection.By this method,the need of automatic and large scale measurement of fiber length is satisfied.The core of this method is the use of skeleton segments instead of discrete pixels as the basic data for the detection of fiber,so as to avoid the influence of complex fiber intersection on the efficiency and accuracy of fiber detection,by which the complex crossing fibers can be automatically separated.Experiment results suggest that the method is reliable and effective.The measured maximum relative error of fiber length between this method and manual method is only 5.2%,and the fiber length distribution curve obtained by the method is very consistent with manual processing results.The 3D SEM reconstruction is applied to the measurement of fiber orientation,and an innovative measurement method of fiber orientation is put forward.Firstly,the epipolar geometry between the stereo pairs of SEM images is analyzed,and it is found that the epipolar lines in the SEM images approach horizontal,and the difference between the heights of the corresponding epipolar lines are fixed.Through detailed analysis and comparison,bilateral filtering and geodesic active contour model are selected to smooth and segment the SEM images respectively.Under the epipolar constraint and the highest-lowest contour point constraint,the matching process of fiber contour is simplified,the 3D profile of fiber is then recovered.In order to reduce the influence of noise and outliers in the 3D profile points,the 3D cylindrical fitting method based on coordinate transformation is improved by using RANSAC algorithm,thus the correct fiber cylinder surface is fitted and the direction of fiber axis is determined.Finally the fiber orientation measurement method based on 3D SEM reconstruction is proposed.By applying the method to the fiber orientation measurement in the fractured surfaces of CF/PLA composite film,the experimental results show that the method performs a high accuracy and can complete the fiber orientation measurement work well.By applying the method to the fiber orientation measurement in the fractured surfaces of CF/PLA composite specimens,the experimental results show that the method can obtain abundant and accurate quantitative analysis results,so that the study of the fiber orientation of short fiber composites by means of SEM is improved from qualitative analysis to quantitative characterization.The research work in this paper provides a practical method for quantitatively characterizing the microstructure of short fiber reinforced composites,which provides objective data support for the design and research of composite materials.