A regression analysis of filler particle content to predict composite wear

It has been hypothesized that composite wear is correlated to filler particle content. There is a paucity of research to substantiate this theory despite numerous projects evaluating the correlation.

The purpose of this study was to determine whether a linear relationship existed between composite wear and filler particle content of 12 composites.

In vivo wear data had been previously collected for the 12 composites and served as basis for this study. Scanning electron microscopy and backscatter electron imaging were combined with digital imaging analysis to develop "profile maps" of the filler particle composition of the composites. These profile maps included eight parameters: (1) total number of filler particles/28742.6 microns2, (2) percent of area occupied by all of the filler particles, (3) mean filler particle size, (4) percent of area occupied by the matrix, (5) percent of area occupied by filler particles, r (radius) 1.0 < or = micron, (6) percent of area occupied by filler particles, r = 1.0 < or = 4.5 microns, (7) percent of area occupied by filler particles, r = 4.5 < or = 10 microns, and (8) percent of area occupied by filler particles, r > 10 microns. Forward stepwise regression analyses were used with composite wear as the dependent variable and the eight parameters as independent variables.

The results revealed a linear relationship between composite wear and the filler particle content.

A mathematical formula was developed to predict composite wear.