Assessment of AASHTO Load-Spreading Method for Buried Culverts and Proposed Improvement

• Published in: Transportation research record Vol. 2675; no. 11; pp. 877 - 887
• Main Author: Katona, Michael G.
• Format: Journal Article
• Language: English
• Published: Los Angeles, CA SAGE Publications 01.11.2021
• ISSN: 0361-1981; 2169-4052
• DOI: 10.1177/03611981211021546

AASHTO’s ad hoc method (AAM) for predicting free-field soil stress under a rectangular loading area is a simple and very useful tool for the analysis of buried culverts subject to vehicular wheel loads. AAM assumes the surface load spreads with soil depth into an ever-increasing rectangular area whose dimensions are controlled by a constant spread angle θ usually taken as 30°, denoted as AAM-30°. Both simplified and comprehensive culvert analysis procedures utilize AAM predictions for adjusting pressure distributions acting on the culvert periphery. Also, AAM-30° is routinely used to determine the two-wheel soil interaction depth, in which the combined effect of both axial wheels need to be considered. To date, a thorough accuracy analysis of AAM-30° has not been published in the open literature. This paper provides a unique and rigorous evaluation of AAM-30° using an exact solution from an elasticity-based model (EBM) of a homogeneous half-space with rectangular surface load. One key discovery is the depth parameter called y*, which is the soil depth at which AAM-30° peak-stress prediction exactly matches the exact EBM solution. Moreover, it is shown that y* may be determined by a simple, yet accurate formula that only depends on the square root of the load area. However, the investigation reveals that AAM-30° significantly underestimates peak stress in the shallow-depth zone 0 < y < ½y* by as much as 31.3% of the applied surface pressure. As this is a large nonconservative error it cannot be ignored. Accordingly, a very simple modification is introduced called AAM-θ*, in which θ* is a spread angle that linearly increases to 30° at soil depth ½y* and thereafter θ* remains constant at 30°. An accuracy evaluation of AAM-θ* reveals an order of magnitude increase in accuracy in which the small residual error is conservative, not nonconservative. The paper concludes with discussions on applying AAM-θ* to the analysis of buried culverts when using either simple or finite element model solution procedures.