A series of truly triaxial tests were performed on two sizes of a railroad ballast under different stress paths and with a constant (initial) field density. The results can permit development of nonlinear elastic and plasticity constitutive models. The commonly used resilient modulus model and a variable moduli model were developed as simplified characterization. The behavior of the ballast was found to be dependent on the state of stress, stress paths and particle size.

Methods of estimating the ultimate bearing capacity of piles in sand from the results of static cone and standard penetration tests are compared with the results of load tests on driven and bored piles of different sizes and embedment ratios in the sand bearing stratum. An empirical reduction factor of the ultimate unit point resistance in sand is established, which decreases with greater pile base diameter and with greater average static cone or standard penetration resistance near the pile point. Similarly, for piles in stiff fissured clay an empirical reduction factor of the undrained shear strength is established, which decreases with greater pile base diameter and is greater for driven than for bored piles. The ultimate unit skin friction of piles in a given sand or clay is practically independent of the pile diameter.

Proper planning of preloading requires time predictions that cannot be obtained reliably from laboratory odometer tests. Time forecasts should be based on field permeability tests or test fills. The settlement versus log time curve of preloading can assume a variety of shapes, depending on the geometry and permeability of sand‐silt contained within the consolidating clay. Experience indicates that three types of curves are recurrent: (1) S‐shaped; (2) bilinear with increase in slope; and (3) linear. The first kind shows that settlement is about completed; for (2) and (3), approximate equations are proposed for predicting the settlement of the permanent structure from settlements observed during preloading. Examples are presented from two construction sites in Greece.

According to current practice, the average tested capacity of single piles is multiplied by a predefined traction in order to obtain the design capacity of a single pile. This fraction remains constant regardless of the number of piles tested. In this paper a simple procedure is developed whereby advantage can be taken of the increased reliability of the data that follows from the increased number of piles tested. It is shown that for the same reliability, increasing the number of tests can substantially increase the fraction by which the average of test results is multiplied in order to obtain the design value. The increase is significant even if the number of tests increases only from one to two. A simple procedure is also developed for taking into account the number of samples tested in determining representative strength of a material in an installation.

A method of solution for quasistatic bending problems of Timoshenko beams on linear viscoelastic foundations obeying Winkler's hypothesis is developed by using the eigenfunctions derived from free vibration problems of Timoshenko beams resting on no foundation and by using the correspondence principle between the viscoelastic problem and the associated elastic problem. Series solutions of the eigenfunctions are rapidly convergent and give accurate results by taking only the first several terms of the series. A successive iteration method for evaluating the eigenvalues is also presented. Numerical results for the variations of deflection and reactive force in space and time are illustrated for a Timoshenko beam with free ends on the viscoelastic foundations of the Kelvin, Maxwell, and Standard linear solid types. The effect of shear deformation on the flexure of beam is discussed.

Tests were conducted to determine the effect of a sloping dredge line on passive earth resistance in front of a bulkhead and bending moments in sheet piling. Results are interpreted relative to moment reduction procedures in current use.