Error Analysis of Non-Destructive Ultrasonic Testing of Glass Fiber-Reinforced Polymer Hull Plates

Glass fiber-reinforced polymer (GFRP) ship structures are generally fabricated by hand lay-up; thus, the environmental factors and worker proficiency influence the fabrication process and presence of error in the non-destructive evaluation results. In this study, the ultrasonic testing of GFRP hull plate prototypes was conducted to investigate the statistical significance of the influences of the design parameters, e.g., the glass fiber weight fraction (Gc) and thickness variations, on the measurement error. The GFRP hull plate prototypes were fitted with E-glass fiber chopped strand mats (40 wt % content) with different thicknesses (7.72 mm, 14.63 mm, and 18.24 mm). The errors in the thickness measurements were investigated by conducting pulse-echo ultrasonic A-scan. The thickness variation resulted in increased error. Furthermore, hull plate burn-off tests were conducted to investigate the fabrication qualities. Defects such as voids did not have a significant influence on the results. The statistical analysis of the measurement errors confirmed that the thickness variations resulted in a strong ultrasonic interference between the hull plates, although the hull plates had similar specific gravity values. Therefore, the ultrasonic interference of the layer group interface should be considered to decrease the GFRP hull NDE errors with respect to an increase in the thickness and Gc.