It is well established that damage to structures built on expansive soils is mainly caused by changes in soil suction. Suction changes are generally attributed to changes in environmental conditions such as change in water table depth, surface irrigation and landscape, resulting in changes in the surface and groundwater regime. Slabs-on-grade must resist both long-term and short-term moisture-change induced soil volume change. The design of residential structures in and regions is especially challenging because the soil experiences large variations in metric suction and associated substantial volume change. As a result, a large number of houses experience minor to severe distress.

Unsaturated soil mechanics theory is used in the determination of unsaturated soil behavior. It is the purpose of this research work to help bridge the gap between theory and practice in the design of residential foundations on expansive soils. One part of this study relates to investigating the depth and degree of wetting associated with moisture flow through expansive soils through modeling and field studies in semi-arid region for typical residential construction development, as well as assessment of foundation performance. A number of steps were taken towards the goal of developing a better understanding of expansive soils behavior and field conditions leading to problems with expansive soils. These steps include: (1) numerical modeling of moisture flow through expansive soils in one- and two-dimensions. Two extreme surface flux conditions were considered, desert and excessively irrigated turf landscapes. The numerical results are applicable to regions with low to moderate expansion potential and Phoenix, Arizona environmental conditions. (2) Development of map illustration to identify locations with low to medium swell potential in the Phoenix Valley. (3) Comparisons of the numerical results to field evidence on depth of wetting and depth of active zone. (4) Evaluation of stability, convergence and numerical challenges for unsaturated moisture flow through expansive soils using Richards' equation. Sources of numerical instabilities were identified and potential improvements discussed. (5) Survey of Arizona region practitioners to identify current design and construction practices, and (6) Analysis of forensic investigations to identify the nature and common causes of residential distress.