Detecting surface waviness prior to painting

Charles Lloyd
Visual Performance, Inc.

Proceedings of the 1999 Coatings and Appearance Testing Conference, Society of Manufacturing Engineers


This paper describes the first of two laboratory evaluations designed to optimize the visibility of large topographical defects (e.g., surface waves, bond line read-through, high and low spots, blisters, and ejector pin marks) on unpainted automotive panels molded from sheet molding composite (SMC). In the first evaluation defect visibility was measured as a function of four practical inspection facility design variables: Viewing Angle, Lamp Blur, Illuminance, and Surface Type. All four of these variables had statistically-reliable effects on ratings of defect visibility made by 10 observers.

An analysis of the optical properties of polish swirl marks in paint suggested ten lighting and viewing parameters will affect defect visibility. This paper describes two designed experiments in which the effects of the first four variables on this list were isolated and measured: 1) viewing angle, 2) defect-light source angle, 3) light source size, and 4) light source blurring. Thirteen observers provided ratings of defect of visibility for 133 combinations of lighting and viewing parameters. Statistically reliable estimates of the effects of the four independent variables were obtained. These results provide objective guidance for the design of an effective inspection facility.