COMPUTER AIDED DESIGN OF RAFT FOUNDATIONS

Abstract

Raft (or Mat) foundations are most commonly used where the soil encountered has low bearing capacity or the soil mass contains compressible lenses which would cause large differential settlements if isolated footings are adopted. Design of a raft foundation requires determination of bending, twisting moments and shear forces induced in the raft due to external loads usually applied through the columns of the super structure it supports. So a Finite Difference solution to the problem of a raft resting on elastic foundation is adopted. Here a raft is divided into square elements and external column loads are distributed as point forces applied at the corner of the square grids (nodal points). Then the differential co-efficients in the raft slab equation and in the boundary conditions are replaced with the corresponding difference coefficients in the nodes of an appropriate mesh. The solution of this system of linear equations leads to values for the deflection of the flat slab in the chosen nodes. The involved computations demand use of computer and therefore, to enable a civil engineer analyse and design a raft foundation a user friendly software package has been developed (FDRAD — Finite Difference Analysis and Design of Raft) using 'C' language and the features of Visual Basic-5. The structural design of the raft foundation is carried out by dividing the raft into column strips along the column line and middle strips adjoining to column strips. The design is based on Limit State method incorporating the provisions of IS code. The package developed is capable of handling rafts loaded by columns transmitting both vertical loads and moments. The package also can handle soil foundation having soft compressible pockets. The package is of interactive type incorporating the relevant knowledge base. The data input by the user is made easy with the screen help messages at appropriate stages. The package is used to solve a few numerical problems and to make a few parametric studies. The results are presented to illustrate some of the features of the package. The results show that (i) A variation in modulus of subgrade reaction value by two to three times results in only about 17% to 25% change in maximum moment values. (ii) Soft or Stiff pockets b