磨损表面的分形特性研究

The Fractal Character of Wear Surface

用结构函数法对试验所得的往复式滑动磨损表面轮廓曲线进行了分形维数的计算,计算结果表明:磨损表面轮廓具有统计自仿射分形特性.表面轮廓仪的采样间隔、评定长度和取样长度不会明显地影响分形维数的大小,但是当评定长度过小时,轮廓曲线的结构函数双对数坐标图会出现离散化现象.这主要是由于统计自仿射分形的统计特性所决定的.取样长度过小时,会引起分形维数测定较大的误差.

Test results of wear surface morphology dimension characterization show that wear surface profile has statistical selfaffine dimension characters. There are no effects of the profilometer sampling spacing and evaluation length on fractal dimension values of surfaces. However, if the evaluation length is too short, the structure function logarithmic value of surface profile is scattering. This is due to the statistic character of statistical selfaffine fractal. Sampling length which takes as filtering process plays an important role in fractal dimension measurement. If the sampling is too short in length, the evaluation of fractal dimension will have larger deviation.