Injection molded polymer test specimens produced within the frame of a
full factor sampling plan were measured using two different kinds of
instrumentations, i.e. a non-contacting Zygo NewView 100 optical inter
ference microscope and a newly developed Rank Taylor Hobson Talyscan p
robe using a contacting stylus with very low forces. Surface roughness
and texture data were subjected to statistical analysis of variance w
ith the purpose of establishing quantitative relationship to processin
g parameters and to the traditionally used visual judgement of produce
d components. The paper will present the experimental setup and the ma
jor results of the data analysis, i.e., a statistical based selection
of amplitude, height distribution, and spatial parameters as well as t
he relative influence of the mold, the material, and the processing pa
rameters to the surface texture. (C) 1998 Elsevier Science Ltd.