Development of an SEM image analysis method to characterize intumescent fire retardant char layer.

• An analysis method for SEM images of intumescent fire retardant char layers is proposed. • Foams in an intumescent fire retardant char layer are characterized quantificationally. • Appropriate addition of CaAlCO 3 -LDHs makes the char layer compact and uniform. • Unsatisfactory result image after color transformation is returned to object segmentation. A new characterization method is proposed to investigate the microscopic morphology of intumescent fire retardant (IFR) char layer observed under a scanning electron microscope (SEM). The method consists of object segmentation extracting the object, color transformation improving the recognition of subtle details, and statistical analysis obtaining the features (area, average, and standard deviation) of the object in the SEM images. There is a descending trend in the standard deviation of the foams with the increase of CaAlCO 3 -LDHs (layered double hydroxides) added to the coatings. Specifically, the number of the foams in the IFR coating with 5 wt% of CaAlCO 3 -LDHs addition is increased from 102 (that of the char layer without CaAlCO 3 -LDHs) to 123. Meanwhile, the standard deviation of the foams in the char layer is decreased from 6383.6 to 4304.5, demonstrating that an appropriate addition of CaAlCO 3 -LDHs can improve the foam quality of the IFR coating by making it more compact and uniform, and enhance the thermal insulation of the char layer covered on the substrate surface. [ABSTRACT FROM AUTHOR]