A simplified model based on the one‐dimensional wave equation, which accounts for the interaction of the pounder and the soil, and the propagation of stress wave in the soil during the dynamic compaction of loose granular soil is presented. The soil beneath the pounder is represented by a nonlinear soil column, while the surrounding soil is represented by a series of springs and dashpots. The spring simulates the dynamic soil stiffness and the dashpot accounts for the radiation damping effect. The input soil parameters of the model can be determined in the laboratory or estimated from correlations with measured field data. In spite of the various simplifying assumptions in the model and the predictive method, computed results, such as pounder penetration and degree and depth of improvement, show an encouraging measure of agreement with available field measurements from two dynamic compaction projects. The proposed model is potentially useful for the analysis of dynamic compaction of loose granular soil.

Two methods found in current engineering practice for the dynamic response analysis of reinforced‐soil retaining‐wali structures are reviewed. The first method is an iterative equivalent linear clastic approach, and the second is an incremental elastic approach. Essential features of these two approaches have been implemented in the finite element programs QUAD4B and TARA‐3. The predictive capability of the two methods of analysis are evaluated by comparing the computed time histories of horizontal accelerations and dynamic forces in the reinforcement layers with field test data previously reported for a full‐scale instrumented wall reinforced with metallic strips. The field test data was obtained by subjecting the instrumented wall to seismic excitations generated by buried explosives detonated with delays. Accelerations on the order of 0.08 g and duration of 0.70 s were generated in the blast series considered for this study. The results of this study show that the incremental elastic approach used in TARA‐3 gives the best prediction of dynamic wall response under blast loading.

A theoretical investigation of the static and dynamic torsional stiffness of arbitrarily shaped rigid foundations embedded in a homogeneous half‐space, using a quadratic‐element‐based boundary element algorithm, is presented. Extensive parametric results are displayed in the form of dimensionless graphs for a variety of basemat shapes and a wide range of embedment depths and types of contact between the foundation sidewalls and the surrounding soil. Based on these results, simple algebraic expressions are also developed, which can be readily used by engineers in practical applications. Discussion of the results provides insight into the nature of torsional response. A numerical example illustrates the practical applicability of the developed formulas, while a companion paper studies the torsional radiation damping of embedded foundations.

Radiation damping of torsionally oscillating, arbitrarily shaped rigid foundations embedded in linear hysteretic homogeneous soil is studied parametrically with a rigorous boundary element formulation. A simplified analytical model, based on sound physical approximations and calibrated against the numerical results, is also developed. The basemat shapes studied include rectangles of aspect ratios up to 6, circles, triangles, and T‐shapes. Particular emphasis is placed on the effects of the depth of embedment and the extent of contact between sidewalls and surrounding soil. It is shown that even foundations placed in an open trench, without sidewall‐soil contact, radiate energy more effectively than surface foundations with an identical basemat. Additional radiation damping is generated from vertical side‐walls, even if they are in good contact with the soil only over a small height compared with the embedment depth; hence, separation or slippage between sidewalls and soil near the ground surface would not substantially affect the amount of torsional damping. Physical arguments are put forward to explain these findings and develop insight to the problem. The paper concludes with an illustrative numerical example.

Changes in tip resistance, local friction, and friction ratio following blasting in saturated, sandy alluvium are reported. Cone penetration tests (CPT) were performed one week prior to blasting and one, three, and 18 weeks after blasting. Blasting initially decreased the tip resistance by 62%, decreased the local friction by 30%, and increased the friction ratio by 100% of the preblast values. From one week to 18 weeks after blasting, the normalized tip resistance increased by 18%, the normalized local friction decreased by 39%, and the normalized friction ratio decreased by 80% of the postblast values. The rate of increase of normalized tip resistance observed in this research is less than that reported from tests conducted in warmer climates but is greater for tests in colder climates. This suggests that the rate of aging of sands may be temperature‐dependent. Time‐dependent changes in local friction observed in this research have not been reported in the literature.

The resonant column technique is used to identify the most significant factors affecting the low‐strain shear modulus of sand cemented with various types of cementing materials. Results show that the increase of low‐strain shear modulus of cemented sand greatly depends on the number of contact points between soil particles at which bonds are developed by the cementing materials. The writers conclude that the effective grain size $D_{10}$ and coefficient of uniformity $C_u$ are the most important index properties of soil controlling the number of interparticle contacts in sand. In addition, the cementation level, which is expressed in terms of percent void filled by cementation, and properties of cementing material also significantly affect the shear modulus of cemented sand. Results indicate that well‐graded sand with small effective grain size can be significantly improved in stiffness by cementation because of its grain number of contact points between soil particles. An empirical equation is developed for quantitatively estimating low‐strain shear modulus of cemented sand.

The influence of seepage on the stability of a sandy slope is investigated. Two failure modes that influence the slope stability can be distinguished. A `continuum' mode if an outward flow is present and a `single particle' mode for an inward flow. A description of these two failure modes is presented and compared with the results of experiments. The description of the continuum mode is based on well known continuum soil mechanics, the single particle mode is based on elaborating the momentum equation for one single sand particle. The single particle mode results in a lower stability than could be expected when using the continuum mode. In the experiments the stability of a sandy slope was tested for different slope angles and hydraulic gradients. In these experiments the difference between the two failure modes was clearly present and the transition between stable and unstable slope was in agreement with theory.

Closed solutions are obtained for the vertical normal stress at any point within an elastic half-space for two loading conditions. In the first case, the stress is caused by a linearly distributed load acting on a superficial polygon. In the second, the stress is produced by uniform distributed load acting on a polygon in the interior of the half-space. The expressions are obtained by integration over a triangle of well-known solutions for stresses induced by concentrated loads: the Boussinesq and Mindlin formulas in the first and second cases respectively. Since any polygonal area can be broken into several triangles, these integrations readily lead to solutions valid for polygonal surfaces with the same type of loading. These expressions can be included in stress estimation computer programs commonly used in soil mechanics to evaluate settlement of superficial and deep foundations instead of approximate numerical methods. Notwithstanding the performance of modern computers, the expressions are considered to be useful from both research and practice point of view, since they lead to important computer-time savings and improved accuracy.

The current procedure for controlling the quality of compaction in the granular shell of an earth dam is to compare the in situ density with the respective maximum laboratory density. This can be tedious on a routine basis during construction. A simple method is described here that does not replace but can readily enrich the results of the current procedure. It consists of taking readings at specified points on the settlements induced by the vibratory roller each time it goes over those points. The induced-settlement readings, correlated with the current results, provide additional data for controlling the attained in situ compaction. They are recorded and evaluated in two distinct forms: (1) the independent; and (2) the cumulative. In the first form, each reading starts fresh from zero, and in the second it is added successively to each new pass of the roller. Readings in the first form are seen primarily as point samples for quality control of compaction. Readings in the second describe the orbit traced by the settlements under repetitive dynamic loading. Finally, 630 results from a specific case are analyzed and commented on with reference to the two forms.

Eight long-term multiple-stage-loading oedometer tests were performed on a fibrous peat from Middleton, Wisconsin. Both conventional-size and large specimens were treated with pore-pressure measurement. Load-increment duration and load-increment ratio were additional variables of the testing program. Time-settlement curves show that the coefficient of secondary compression ($C_{\alpha }$) is not constant but increases in time under constant effective stress. Plots of $C_{\alpha }$ versus compression index ($C_c$) indicate that, for this peat, a poor linear relationship exists between $C_{\alpha }$ and $C_c$, and this relationship worsens if tertiary compression points are included. Plots of $C_{\alpha }/C_c$ versus void ratio show significant scatter but indicate that, in general, the $C_{\alpha }/C_c$ ratio increases with decreasing void ratio. Load-increment ratio and sample disturbance are significant factors affecting the $C_{\alpha }$ versus $C_c$ relationship. The fact that inorganic soils do not exhibit exaggerated creep effects or tertiary compression may explain, in part, the success of the $C_{\alpha }/C_c$ concept for clays.

A new and accurate formulation of the stiffness matrix for an axisymmetric interface or joint element is presented. This element is formulated to model rock joints and soil-soil and soil-structure interfaces in the finite element analysis of axisymmetrical solids of revolution. The formulation is done in the local coordinate system and the closed-form solution of the interface stiffness matrix is obtained. The advantages of the new formulation are: (1) It is exact and accurate under all conditions, versus the various approximate forms; (2) it can be easily used in a finite element code; and (3) the planar interface element is a special case of the new interface element. As such, the stiffness matrices of the planar and axisymmetric interface elements can now be expressed in a unified form, thus eliminating the need for two different types of interface elements. An example problem involving a rigid footing is used to show the advantage of using the new formulation in modeling axisymmetric interface behavior.

The purpose of this study is to investigate the relationship between swell and degree of saturation for a compacted clay. Laboratory tests indicate that the percent swell increases in direct proportion to the degree of saturation at the end of the test. The linear relationship does not intercept the initial compacted degree of saturation. The coefficient of swell can be affected by the compacted effort. The dry, hard clods are broken down by the higher compactive effort. This reduction in interclod voids probably causes the significantly different coefficient of swell. Swell tests compacted to low, dry densities have a final saturation degree below 90%, even after two cycles of secondary swell. This indicates the presence of entrapped air. The main factor governing entrapped air for this soil appears to be the initial compacted void ratio. By itself, the degree of saturation does not seem to reliably indicate future potential for swell. One problem is the presence of entrapped air, which can result in a saturation degree below 90%, even when the sample has essentially stopped swelling.