PREDICTION OF SOIL REINFORCEMENT LOADS IN MECHANICALLY STABILIZED EARTH (MSE) WALLS

Proper estimation of soil reinforcement loads and strains is key to accurate design of the internal stability of geosynthetic and steel reinforced soil structures. Current design methodologies use limit equilibrium concepts to estimate reinforcement loads for internal stability design, with empirical modifications to match the prediction to observed reinforcement loads at working stresses. This approach has worked reasonably well for steel reinforced walls but appears to seriously overestimate loads for geosynthetic walls. A large database of full-scale geosynthetic walls (16 fully instrumented, full-scale geosynthetic walls and 14 walls with limited measurements) and 20 fully instrumented, full-scale steel reinforced mechanically stabilized earth (MSE) wall sections was utilized to develop a new design methodology based on working stress principles, termed the K sub 0 Stiffness Method. This new methodology considers the stiffness of the various wall components and their influence on reinforcement loads. Results of simple statistical analyses to evaluate the ratio of predicted to measured peak reinforcement loads in geosynthetic walls were telling: the AASHTO Simplified Method results in an average ratio of predicted to measured loads of 2.9 with a coefficient of variation (COV) of 86%, whereas the proposed method results in an average of 1.12 and a COV of 41%. The proposed method remains accurate up until the point at which the soil begins to fail (approximately 3 to 5% strain). For steel reinforced MSE walls the improvement was more modest: AASHTO's Simplified Method results in an average ratio of predicted to measured loads of 1.04 with a COV of 51%, whereas the new K sub 0 Stiffness Method results in an average of 1.12 and a COV of 35%. The objective of the method is to design the wall reinforcement so that the soil within the wall backfill will not reach a state of failure consistent with the notion of working stress conditions. This soil failure limit state is not considered in the design methods currently available, yet, given the research results presented herein, is likely to be a controlling limit state for geosynthetic structures. The fruit of this research is a more accurate method for estimating reinforcement loads, thereby reducing reinforcement needs and improving the economy of MSE walls. The scope of this research was limited to MSE walls that utilize granular (non-cohesive, relatively low silt content) backfill.