Abstract A laboratory study was undertaken to investigate the hydraulic conditions under which stones in a model riprap apron at the outlet of circular pipes would move and thereby initiate apron instability. Cameras deployed both above and below the water surface acquired video records of possible stone motion at the culvert outlet, thus allowing reliable assessment of eventual apron instability. A conceptual model is proposed that emphasizes the qualitative difference between low-tailwater and high-tailwater cases for the modeling of apron instability. It differs from the earlier model in its choice of dimensionless variables and hence in the criterion for the different regimes. The choice of coordinates allowed asymptotic arguments that led to relatively simple expressions for the delineation of stable regions. Comparison of the present data with two widely used models and of the proposed stability regions with available data indicates that the largest practical quantitative differences in prediction will occur for the high-tailwater case, being generally more conservative than one, but less generally conservative than the other. Even in the low-tailwater cases, however, the proposed model should be less prone to misidentification of unstable conditions.
Lyn et al. (Thu,) studied this question.