Role of suction stress on service state behavior of geosynthetic-reinforced soil structures

Several design guidelines adopt limit state methods (e.g., earth pressure or limit equilibrium slope stability methods) to estimate the reinforcement loads for geosynthetic-reinforced soil structures (GRSSs). Field measurements usually reveal lower tensile loads in the reinforcements than that predicted by conventional design methods. Consequently, limit state methods have been criticized for being conservative or inaccurate. However, these lower-than-expected loads are primarily due to redundancy in design, attributed to several factors such as toe resistance, soil volumetric dilation, underestimation of soil shear strength, and the effect of suction stress. While these factors commonly contribute to the performance of the GRSSs, they are not accounted for in design procedures. This disregard is accredited to complexities and uncertainties associated with reliable quantification of these factors during the life span of the structure. By properly quantifying the role of suction stress, this study aims to quantitatively explain a part of the discrepancy which exists between the tensile loads in the reinforcement, which are estimated by working-condition design methods versus those by limit state design methods. A suction stress-based formulation is presented for calculating the active earth pressures coefficient (Ka) of unsaturated backfill under unsaturated steady flow conditions. The formulation is derived by implementing an analytical solution of one-dimensional steady flow into a limit equilibrium-based effective stress analysis. The proposed formulation is used in conjunction with the earth pressure method to illustrate the effect of suction stress on predicted reinforcement loads. Two backfills, referred to as marginal and high quality backfill, along with three representative annual rain events are examined. The results are compared with two classical earth pressure methods as well as two empirical methods. The results show that the suction stress contribution can lead to significantly lower reinforcement loads than that predicted by classic earth pressure methods for backfill types. Empirical methods lead to lower loads in some cases and higher predicted loads in others. The proposed formulation provides a mechanics-based approach to explain a part of the discrepancy which exists between the measured tensile loads in the reinforcement in the field versus those by limit state design methods.