Development of Surface Roughness Generation Model for CFRTP Manufactured by LFT-D

Abstract

In this study, we propose a new surface generation model for carbon fiber reinforced thermoplastics (CFRTP) manufactured by the long fiber thermoplastic-direct (LFT-D) method. CFRTP are considered to be a next-generation structural material because of their high productivity as well as high mechanical strength and lightness. Conversely, CFRTP have a rough surface, which does not meet the automotive outer panel standard of a “class A surface.” In the present study, we establish a surface roughness generation model based on a thermal shrinkage mismatch of thermoplastic resin to carbon fiber and non-uniform carbon fiber distribution. Furthermore, we construct a surface roughness estimation formula based on the model. In the calculation, a cross-sectional image of CFRTP is divided into many vertical segments. Subsequently, the thermal shrinkage of each segment is calculated with a standard deviation, an average, and a probability density of the amount of carbon fiber in each segment. The surface roughness of the manufactured CFRTP was measured using a surface profilometer. The result showed that the arithmetic surface roughness increased with the volume fraction of carbon fiber. We applied the surface roughness calculation to cross-sectional images of the specimens. Consequently, the estimated surface roughness showed the same tendency, in which the surface roughness increased with the volume fraction of carbon fiber. The slope of a regression line of the estimated surface roughness with respect to the volume fraction was 0.010, which was almost the same (0.011) as the slope of a regression line of the measured surface roughness. Furthermore, the estimation formula using a thermal shrinkage effective depth of 395 μm was able to estimate the surface roughness within a 3% average error. Using the estimation formula, it was predicted that the surface roughness increased with the standard deviation of the amount of carbon fiber in a segment. To confirm the reliability of the model and the formula, we measured the standard deviation of the amount of carbon fiber in CFRTP specimens, showing that the trend for CFRTP specimens matched the estimated values.