Settlement of Shallow Foundations on Clay—A Database Study: Fid-Bau Portal

Settlement of Shallow Foundations on Clay—A Database Study

Bahmani, M. / Briaud, J.-L.

Generally, the first choice in the selection of a foundation for buildings and other structures is a shallow foundation because it is more economical than a deep foundation and easier to construct. However, shallow foundations are usually considered to settle more than deep foundations. This is a fallacy as well-designed shallow foundations do not settle more than well designed deep foundations. This research aims to give guidance on selecting the soil modulus to calculate the long-term settlement of shallow foundations on clay, given a chosen probability that the calculated settlement will be larger than the measured settlement. The settlement of shallow foundations depends on the properties of the soil, the load on the soil, and the geometry of the foundation. The soil properties vary from one site to another, while the geometry of the foundation can be selected to keep the settlement due to the load within tolerable limits. In some cases, the settlement of the foundation is recorded along with the soil properties, and the data are published in the open literature. Over time, the number of such cases has increased and, if well-organized in a database, can be very useful in practice. Indeed, comparing the measured data with the predicted data can guide the practitioner in the safety margin to be used for a given prediction method. In this paper, a database of shallow foundation settlement records on clay called “TAMU-SHAL-CLAY” has been created by collecting case records from the open literature and organizing them in a spreadsheet available free of charge. The measured settlements are compared with the predicted settlements using the elastic equation while using a long-term soil modulus estimated as a multiple of the undrained shear strength of the clay. In addition to the predicted versus measured settlement comparisons, the probability that the predicted settlement will be larger than the measured settlement is presented as a prediction multiplier function. The results can help the practitioner decide at which level of probability this practitioner wishes to operate.