D4 Fractography Microscopy

Topic

This model analyzes the fracture surface resulting from a laminate failure to identify the dominant mechanism responsible for it. Because this analysis is conducted after the failure occurs rather than during the process, it is termed a "post-mortem" analysis, performed using scanning electron microscopy on the fracture surface itself.

Scanning electron microscopy of the fracture surface identifies which mechanism dominated the laminate failure: fiber pull-out, where the fiber is extracted from the matrix, leaving an empty void; debonding, where the fiber separates cleanly from the matrix without breaking; fiber breakage, where the fiber itself snaps, leaving a flat fracture surface; or crack deflection, where the crack diverts around the fibers rather than cutting straight through them. Since each mechanism leaves a distinct mark on the fracture surface, microscopy allows for the identification of the dominant failure mechanism simply by observing that surface.

When fiber pull-out is identified as the dominant mechanism, the model also measures the average pull-out length on the fracture surface—that is, the length of the fiber exposed after extraction from the matrix, prior to the fiber breaking instead of continuing to slide out. If the measured average pull-out length matches the theoretical critical fiber length—the length below which a fiber pulls out rather than breaking, a value established in a separate model within this suite describing the fiber-matrix interface—then the theoretical critical length is experimentally verified; conversely, a significant discrepancy between the two lengths indicates that the other model requires revision. Stereological counting, for its part, determines the fraction of the total fracture surface area corresponding to each of these mechanisms: by superimposing a grid onto the fracture surface and noting which mechanism coincides with each grid point, one obtains the proportion of the fractured area attributable to pull-out, debonding, fiber breakage, and crack deflection. Because each grid point is tallied according to the coinciding mechanism, the sum of these counts directly yields the area fraction fractured by each mode. For this counting process, the model reuses the same stereological method already employed—for the same purpose—by the equivalent model for metals.

Consequently, this model—along with the fractographic analysis with which it began—provides results that serve to cross-verify two other models in the set: the measured mean pull-out length, when compared to the theoretical critical length, verifies the model describing the fiber-matrix interface; and the area fraction fractured by each mechanism, determined via stereological counting, verifies the model describing the laminate's toughening mechanisms.

ID:4425