The St. Luke's Medical Tower is the Texas Medical Center's tallest building and Houston's tallest building to open the 1990s. The combination of the following unique foundation features reduced development costs by over $1,000,000: (1) tallest soil-supported building on shallow foundations in the Southwest; (2) temporary dewatering system designed to function as the permanent system; (3) excavation-bracing system designed to form the permanent basement wall and only individual braces were temporary; (4) basement walls esigned to accept loads from future contiguous towers; (5) drilled pier soldier piles installed with polymer drilling fluid (the first use on a major Texas project); and (6) drill pier soldier piles installed in accordance with the new American Concrete Institute Standard Specification for the Construction of Drilled Piers (ACI 336.1-89), the first known use of the specification. The factored load condition was considered fictional in foundation, and basement wall design in that factored load-concrete-subgrade compatibility was not achieved. Significant cost savings was achieved by allowing the geotechnical engineer to be part of the design team beginning with project concept studies and extending throughout underground construction. The geotechnical engineer and the team developed feasible foundation schemes that could be integrated into construction needs, instead of relying only on specialty design builders.

Two mat foundations supporting buildings were analyzed using traditional methods for modeling soil subgrades, as well as more recently developed methods. The primary purpose of the analysis was to evaluate the relative and absolute accuracy of subgrade models that can be used in routine practice. The results indicate that some of the newer methods consistently provide significantly better agreement between calculated and observed behavior compared to the traditional methods. In addition, determination of the appropriate subgrade parameters is more rational with the newer methods. With current computer analysis capabilities, there is no reason to continue use of traditional methods that were reasonable alternatives when only manual calculations could be performed. Detailed recommendations for modeling subgrades in practice are presented, with consideration given to the capabilities of commercially available structural analysis computer software. Other factors that influence mat behavior, such as superstructure interaction effects, are also discussed.

Describes three innovative mat foundation designs and the close interaction required between the structural engineer and the geotechnical engineer. The significance of load deformation prediction reliability in the three different soil profiles is illustrated. The cases reviewed include a three-story office building with single basement build on a mat over peat; a 26-story apartment building with basement built on a modified mat in a thin dense sand stratum over soft clay; and a 19-story hotel with two basements built on a mat in a sand layer over medium clay. The mat of the 19-story hotel was supplemented with selective high capacity piles at the column locations designed to ultimate soil capacity at working loads and utilized to reduce both mat settlement and design mat thickness. The instrumentation used to confirm design assumptions in the three cases is briefly described.

Often, an apparently compatible relationship between mat and soil deteriorates due to the plague of construction details and the design-construction relationship. This paper reviews subgrade reaction in case studies of four landmark buildings in Houston. Concepts related to mat foundation analysis using the finite element method are discussed to acquaint the practitioner with the related soil-structure interaction concepts. Also included is an examination of structural considerations in connection with mat foundation design.

Describes the design of a load-compensated mat foundation on highly compressible soil. The mat was used to support over 800,000 square feet of variable height building. While the design of the mat was mostly routine, the behavior at the mat edges was difficult to determine. The deformations at mat edges were the major concern since they were influenced by the need to raise grades around the building perimeter. The design procedure incorporated soil-structure interaction analysis to determine the extent of lightweight fill zones required to control edge deformations. Settlement monitoring over a period of two years has confirmed the design approach.