CASE STUDY ON THE ESTIMATION OF SUBGRADE REACTION FOR THE DESIGN OF RAFT FOUNDATION ON AEOLIAN SOIL Omar Hamza 1 , Abdul-Hakim Mawas 2 , Alaa Kourdey 3 1 College of Engineering and Technology, University of Derby, Derby DE22 3AW, UK (o.hamza@derby.ac.uk) 2 Structural Engineering Consultant, Dammam, KSA 3 Geotechnical Engineering Consultant, Dammam, KSA Abstract Subgrade reaction is an approximate representation of load-displacement behaviour of the supporting soil. Reliable estimation of the modulus of subgrade reaction (k) is essential for the design of large and multi-column foundations when flexible analysis is used to achieve economical and safe design. This paper presents a case study on the estimation of the modulus of subgrade reaction (k) for the design of different sizes of raft foundations of multi- story buildings constructed on Aeolian soil in Dammam, KSA. A site-specific ground investigation was conducted for this project where boreholes and laboratory classification testing indicated that the soil is mainly loose to medium dens Silty Sand. Plate Load Tests (PLT) were carried out at the foundation level and used to estimates the modulus of subgrade reaction. The load-settlement response from five plate load tests (600 mm diameter) was back analysed using 3D numerical modelling to verify the ground properties. Then, the validated ground model was used to numerically simulate the raft foundations and estimate the k-values. As the size of the foundations is larger than the plat, the modulus of subgrade reaction (k) decreased. The results also showed that the modulus of subgrade reaction obtained from Terzaghi's equation (1955) for the prototype foundation is not always smaller than the k-value obtained from the numerical analysis. The paper presents simple procedures that can be useful for structural and geotechnical engineers who are involved with raft foundation design on sandy soils. Keywords: Subgrade reaction, Raft foundation, Aeolian soil, Plate Load Test, Numerical Analysis 1. Introduction In the design analysis of substructures (flexible raft foundation) or flexible pavement slab supported by the ground, the soil interaction and tendency to deform are commonly represented by a parameter called coefficient or modulus of subgrade reaction (k). The current practice in the selection of subgrade reaction tends to vary, depending upon the availability of laboratory and field testing data as well as the analytical tools adopted in the design. The local tradition and practice and the experience of the structural and geotechnical designers might also influence the choice of the method selected for evaluating the k-value. The earliest method was originally developed by Winkler (1867) assuming the soil medium as a system of identical but mutually independent, closely spaced, discrete and linearly elastic springs (Marto et al. 2012). Accordingly, the modulus of subgrade reaction, k, can be defined as follow: k = p/ (1) Equation (1) defines k-value as the ratio of contact pressure (p) at any given point to settlement ( ) produced by load application at that point. The unit of k-value is typically MN/m 3 or kN/m 3 . Figure 1 shows the analytical model of concrete slab based on Winkler spring assumptions. The common method to estimate the (vertical) modulus of subgrade reaction for the design of shallow foundations is Plate Load Test (PLT) which is typically carried out using 300-600 mm diameter circular rigid steel plate centrally loaded by vertical force (BS1377- Part 9, 1990). The output of PLT is a relationship between the plate pressure (p) and average measured deflection across the plate () as shown in Figure 2. Accordingly, the modulus