Seafloor sediment density flows are the primary mechanism for transporting sediment to the deep sea. These flows are important because they pose a hazard to seafloor infrastructure and deposit the largest sediment accumulations on Earth. The cohesive sediment content of a flow (i.e., clay) is an important control on its rheological state (e.g., turbulent or laminar); however, how clay becomes incorporated into a flow is poorly understood. One mechanism is by the abrasion of (clay-rich) mud clasts. Such clasts are common in deep-water deposits, often thought to have traveled over large (more than tens of kilometers) distances. These long travel distances are at odds with previous experimental work that suggests that mud clasts should disintegrate rapidly through abrasion. To address this apparent contradiction, we conduct laboratory experiments using a counter rotating annular flume to simulate clast transport in sediment density flows. We find that as clay clasts roll along a sandy floor, surficial armoring develops and reduces clast abrasion and thus enhances travel distance. For the first time we show armoring to be a process of renewal and replenishment, rather than forming a permanent layer. As armoring reduces the rate of clast abrasion, it delays the release of clay into the parent flow, which can therefore delay flow transformation from turbidity current to debris flow. We conclude that armored mud clasts can form only within a sandy turbidity current; hence where armored clasts are found in debrite deposits, the parent flow must have undergone flow transformation farther up slope.