|
With the shrinking of global forest resources, the increasing scarcity of timber resources, and a Shaolin countries especially China. How to effectively improve the utilization of timber resources and value in use, today's domestic and international timber industry the researchers attention and research hotspot. Research and Application of the high-performance, high-value-added feature wood composites provide a new direction for this problem. This paper studies the chopped carbon fiber composite wooden function (SCFRW), carbon fiber and wood fiber composite, the composite material has a strong mechanical properties, while giving its good conductivity and electromagnetic shielding performance. The not continuous carbon fibers with a matrix material for the preparation process of the composite material, mixed degree of homogenization seriously affect the macroscopic properties of the problem, based on digital image processing technology, the model SCFRW homogenization. First, the physical performance test template obtained by the experimental system, including electrical conductivity (surface resistivity), mechanical properties (density, internal bond strength, bending strength, elastic modulus, and absorbent swelling), electromagnetic shielding performance, and test data analysis and grasp the variation of chopped carbon fiber wood composite material properties under different preparation conditions, affect the characteristics of composite microstructure uniformity degree of macroeconomic performance; then microscopic image acquisition, test panel, and select 150 amplitude magnification of the same and having the characteristics of the typical shape of the microscopic image, image processing. Image preprocessing based on the combination of the maximum variance method and mathematical morphology method for image segmentation, the binary image of carbon fiber for the segmentation of the target area. Select can effectively reflect the template uniform shape features (area ratio, the average width and aspect ratio, etc.) of the degree of characteristic value as a feature amount extraction; Finally, the minimum width of the region of a single small strip, the average area ratio, average long aspect ratio, the total area of ??the bar area as inputs to the template surface resistivity, elastic modulus, bending strength as output build SCFRW Homogenization BP neural network model. And in the MATLAB environment to verify the validity and reliability of the model. Through this study, the micro-structural features and macroscopic properties of the composite combine, implemented according to the degree of mixing uniformity SCFRW in carbon fiber and wood fiber directly predict the performance of the macro characterization purposes, which provide a favorable basis for the process design of the system board. Based on this, the preparation and adaptive use of composite carbon fiber wood functionality to provide the necessary basic conditions for scientific guidance.
|