Common behavioral trends and characteristics of alluvial channels including rivers are extensively discussed in the literature. However, little is known about the hydrodynamic effects of seepage on alluvial channels. Factors like sand bed resistance, stream power of the channel, incipient motion of bed particles, and geometry of the channel cross section are significantly affected by seepage. This thesis presents the experimental investigations that are aimed to find out the quantitative effect of seepage, through a sand bed in downward (suction) direction, on the above mentioned factors. The problem in the sediment transport analysis is that the knowledge of complex interaction of several parameters with seepage cannot be fully obtained. In order to generalize the results, experiments are conducted in four rectangular smooth walled sand bed flumes under steady and fairly uniform flow conditions. Among the four, one is the Large Tilting Flume (LTF), which is 25 m long, 1.80 m wide and 1.00 m deep and with a seepage length of up to 20 m. This LTF is specially built at hydraulic laboratory, Indian Institute of Science exclusively for the present research work especially on alluvial channel regime. The experimental channels are designed to apply controlled amounts of uniform seepage flow in either direction in one flume (Flume-2), and only in downward direction to other three flumes (Flumes 1, 3 and 4). The application of seepage is perpendicular to the sand bed thickness over a sufficient length of the main channel. Appropriate instruments are used to accurately measure the basic experimental variables such as discharge in the main channel, seepage discharge, flow depth, water surface and bed slope, seepage gradients and cross sectional profiles. Experiments are carried out at different conditions (plane beds and curved shape channels) as explained below: Plane sediment beds Series – 1: Experiments to determine incipient motion of bed particles under no-seepage condition. Series – 2: Experiments on the non-transporting condition of the bed particles under both no-seepage and seepage condition. Series – 3: Experiments on the transporting condition of the bed particles under both no-seepage and seepage condition. Curved shape channels Series – A: Experiments to verify Lane’s (1953) geometric profile against higher discharge than prescribed by Lane (1953). Series – B: Experiments to verify the Lane’s (1953) geometric profile by allowing the discharge prescribed by Lane (1953). Series – C: Experiments to predict the final geometric profile by applying suction to Series – B experiments. A wide range of sediment particles are tested. Five different sized uniform sands (d50 = 1 mm, 0.56 mm, 0.65 mm, 1.00 mm and 1.77 mm) and gravel of size d50 = 8.00 mm are used for experimentation. Among the six sizes, three sizes (d50 = 0.56 mm, 0.65 mm, 2 mm) are used for seepage experimentation. The experimental data from the present experiments along with the available data from other sources on more sizes of sand are analyzed, thus covering a wide range of sand sizes. The following important results are obtained from the analysis. A new resistance equation has been developed for plane sediment beds (with little or no-transport) such that the average velocity in the channel depends on the shear velocity Reynolds number. A careful study has been done on incipient motion and concluded that incipient motion is better explained by critical stream power criterion for plane sediment beds. With the help of critical stream power criterion, a straightforward design procedure using design tables/design curves and analytical methods are presented to solve six possible design problems. For plane and non-transporting beds, in general, the stream power in the channel increases with suction and decreases with injection. The increase and decrease depend on the seepage power intensity parameter (NP), initial value of stream power (Ωbo), and critical stream power of the particles under no-seepage condition (Ωco). An expression relating all the influencing parameters is established to quantitatively estimate the stream power (Ωbs) variation with both the types of seepages, i.e., with suction and injection. It is found that the seepage has a significant influence in changing critical stream power for incipient motion of the bed material and the value is significantly different from the no-seepage critical value. An expression is established to quantitatively estimate the critical stream power with seepage (upward and downward) for a given critical stream power (Ωco) of the bed material under no-seepage conditions and initial stream power (Ωbo). It has been established that critical stream power curve used to define incipient motion is valid only for no-seepage condition of the bed and it cannot be used for sand beds under seepage condition, as seepage effects significantly alters the stream power. From the wide range of experimental data (including the observations from LTF) it is found that suction (downward seepage through the sand bed) enhances the transport or aids the incipient motion of bed particles which are initially at rest. Thus, suction reduces the stability and increases the erosion of bed particles when compared to no-seepage conditions. However, it is found that injection (upward seepage) affects in an opposite way, i.e., it can reduce the transport rate or even inhibit the incipient motion. Thus, injection increases the stability and reduces the erosion of bed particles when compared to no-seepage conditions. Therefore, it is concluded that suction increases the mobility of sand particles where as injection decreases their mobility. An expression to find the incipient motion with seepage (both suction and injection) is established in terms of stream power’s (Ωco, Ωbo and Ωbs) based on the present experimental data along with others' data. With the help of these expressions design procedure is developed for ten types of possible problems. A numerical model for spatially varied flow has been developed with the help of the seepage governing equations, developed in this thesis, to compute flow profiles along the channel length. A methodology of predicting the location of incipient motion section in sand bed channels affected by seepage is also presented. Channel geometry affected by seepage (suction) is established in the form of regression relationships for perimeter, flow depth and slope of the channel. Different combinations (bi-variate and tri-variate) of dimensional and non-dimensional regression relationships are developed. An approach to channel design has been developed based on the application of functional analysis of the salient variables that control the channel behavior. And also, it has been established that, Lane’s (1953) profile almost matches with experimental profile for no-seepage condition. The present investigation clearly shows the significance of seepage in altering the hydraulic and sediment transport behavior of sand bed channels. From the practicing engineer’s point of view it is hoped that present design procedures will be helpful in safe guarding the seepage affected channels.

Doctor of Philosophy(PhD) / Soil Stiffness is an important parameter for any geotechnical engineering design. In laboratory tests it can be derived from stress-strain curves or from dynamic measurement based on wave propagation theory. The second method is a more accurate and direct method for measuring stiffness at very small strains. Until now dynamic measurements have usually been obtained manually from the triaxial test. Attempts have been made to automate the procedure but have apparently failed due to the high level of variability in dynamic measurements. Moreover, triaxial tests of soil can be very lengthy and manual dynamic measurements can be very tedious and impractical for long stress-path tests. In this research a computer program has been developed to automate the stiffness measurement (using bender elements) based on the cross- correlation technique. In this method the program records all the peaks and corresponding arrival times in the cross-correlation signal during the test. The stiffness is calculated and displayed on the screen continuously. The Bender Element enabled to get the small strain shear modulus. An arbitrary “Chirp” waveform of 4 kHz frequency was used for this purpose. Subsequently Bender Element test results were checked by ‘Sine’ waveforms of frequencies 5kHz to 20kHz, as well as by manual inspection of the arrival time. This thesis discusses the method and some of the difficulties in truly automating the process. Finally some results from a number of stress path tests on uncemented and cemented calcareous sediments are presented. Bender elements have been used by many researchers to determine the shear modulus at small strain. Most previous studies have used visual observation of arrival time, which is time consuming and often requires some judgement from the operator. This thesis will describe the use of cross-correlation as a method for automation of Gmax measurement. Cross-correlation has been claimed to be unreliable in the past. However, it will be shown that provided several peaks in the cross-correlation signal are monitored it is possible to follow the variation of Gmax throughout consolidation and shearing. The measurement can be made at regular intervals within the software controlling a stress-path apparatus. Details of the apparatus used and practical considerations including selection of waveform and frequency are discussed. A series of drained cyclic triaxial tests was carried out on artificially cemented and uncemented calcareous soil of dry unit weights 13, 15, and 17 kN/m3 and sheared with constant effective confining stress 300 kPa. Gypsum cement contents of 10%, 20% and 30% of the dry soil weight were used. In addition a series of stress path tests were performed on Toyuora sand samples. Results will be presented for two uncemented and one cemented sand. In addition to the bender elements, all tests had internal instrumentation to monitor axial and lateral strains. Results will be presented for Toyura sand to show that the measurements are consistent with those obtained by other methods. Results will also be presented for carbonate sand subjected to a wide range of stress paths. Finally, results will be presented for the carbonate sand cemented with gypsum. The degradation of Gmax of the cemented soil subjected to variety of monotonic and cyclic stress-paths is presented. Analysis of the results includes assessment of the factors influencing Gmax for uncemented sand. Preliminary analysis indicates that in order of importance these are the mean effective stress, the stress history, void ratio and stress ratio. For cemented sand, Gmax is initially constant and independent of stress path. After yielding the modulus degrades, becoming increasingly stress level dependent and eventually approaches the value for uncemented sand. Factors influencing the rate of degradation are discussed. For the Toyuora sand samples the effects of end restraint on the stress-strain response at small strains were investigated. The conventional method of mounting triaxial specimen has the effect of introducing friction between sample and end platen during a compression test. This inevitably restricts free lateral movement of the specimen ends. Frictional restraint at the sample ends causes the formation of 'dead zones' adjacent to the platens, resulting in non-uniform distribution of stress and strain (and of pore pressure if undrained). On the other hand the specimen with 'free' ends maintain an approximate cylindrical shape instead of barrelling when subjected to compression, resulting in a more uniform stress distribution.

Doctor of Philosophy(PhD) / Soil Stiffness is an important parameter for any geotechnical engineering design. In laboratory tests it can be derived from stress-strain curves or from dynamic measurement based on wave propagation theory. The second method is a more accurate and direct method for measuring stiffness at very small strains. Until now dynamic measurements have usually been obtained manually from the triaxial test. Attempts have been made to automate the procedure but have apparently failed due to the high level of variability in dynamic measurements. Moreover, triaxial tests of soil can be very lengthy and manual dynamic measurements can be very tedious and impractical for long stress-path tests. In this research a computer program has been developed to automate the stiffness measurement (using bender elements) based on the cross- correlation technique. In this method the program records all the peaks and corresponding arrival times in the cross-correlation signal during the test. The stiffness is calculated and displayed on the screen continuously. The Bender Element enabled to get the small strain shear modulus. An arbitrary “Chirp” waveform of 4 kHz frequency was used for this purpose. Subsequently Bender Element test results were checked by ‘Sine’ waveforms of frequencies 5kHz to 20kHz, as well as by manual inspection of the arrival time. This thesis discusses the method and some of the difficulties in truly automating the process. Finally some results from a number of stress path tests on uncemented and cemented calcareous sediments are presented. Bender elements have been used by many researchers to determine the shear modulus at small strain. Most previous studies have used visual observation of arrival time, which is time consuming and often requires some judgement from the operator. This thesis will describe the use of cross-correlation as a method for automation of Gmax measurement. Cross-correlation has been claimed to be unreliable in the past. However, it will be shown that provided several peaks in the cross-correlation signal are monitored it is possible to follow the variation of Gmax throughout consolidation and shearing. The measurement can be made at regular intervals within the software controlling a stress-path apparatus. Details of the apparatus used and practical considerations including selection of waveform and frequency are discussed. A series of drained cyclic triaxial tests was carried out on artificially cemented and uncemented calcareous soil of dry unit weights 13, 15, and 17 kN/m3 and sheared with constant effective confining stress 300 kPa. Gypsum cement contents of 10%, 20% and 30% of the dry soil weight were used. In addition a series of stress path tests were performed on Toyuora sand samples. Results will be presented for two uncemented and one cemented sand. In addition to the bender elements, all tests had internal instrumentation to monitor axial and lateral strains. Results will be presented for Toyura sand to show that the measurements are consistent with those obtained by other methods. Results will also be presented for carbonate sand subjected to a wide range of stress paths. Finally, results will be presented for the carbonate sand cemented with gypsum. The degradation of Gmax of the cemented soil subjected to variety of monotonic and cyclic stress-paths is presented. Analysis of the results includes assessment of the factors influencing Gmax for uncemented sand. Preliminary analysis indicates that in order of importance these are the mean effective stress, the stress history, void ratio and stress ratio. For cemented sand, Gmax is initially constant and independent of stress path. After yielding the modulus degrades, becoming increasingly stress level dependent and eventually approaches the value for uncemented sand. Factors influencing the rate of degradation are discussed. For the Toyuora sand samples the effects of end restraint on the stress-strain response at small strains were investigated. The conventional method of mounting triaxial specimen has the effect of introducing friction between sample and end platen during a compression test. This inevitably restricts free lateral movement of the specimen ends. Frictional restraint at the sample ends causes the formation of 'dead zones' adjacent to the platens, resulting in non-uniform distribution of stress and strain (and of pore pressure if undrained). On the other hand the specimen with 'free' ends maintain an approximate cylindrical shape instead of barrelling when subjected to compression, resulting in a more uniform stress distribution.

Understanding the mechanisms and adaptations that allow only certain species to thrive in the potentially stressful foredune environment requires a knowledge of the basic ecophysiology of foredune species. Ecophysiological measurements were conducted on the foredune pioneer species Arctotheca populifolia (Berg.) Norl., Ipomoea pes-caprae(L.) R. Br. and Scaevola plumieri (L.) Vahl. and showed significant differences among species with respect to the physiology associated with biomass production, water and nutrient relations. Differences related to CO₂ assimilation included differences in photosynthetic and respiratory rates, susceptibility to light stress and leaf and stem non-structural carbohydrate concentrations. These resulted in differences in primary production rates of shoots. Mechanisms leading to the differences in CO₂ assimilation among species included differences in stomatal behaviour, carboxylation efficiencies, efficiencies of utilisation of incident photosynthetic photon flux density (PPFD) and rates of ribulose-1,6-bisphosphate (RuBP) regeneration. Correlated with differences in photosynthetic capacity were differences in chlorophyll contents but not differences in leaf nitrogen content. Differences in interspecific stomatal behaviour resulted in significantly different transpiration rates which in combination with differences in assimilation rates resulted in differences in water-use efficiency. The absolute amounts of water transpired, although significantly different among species, were moderate to high in comparison with species from other ecosystems and were typical of mesophytes. Transpiration rates in combination with plant hydraulic conductances and soil water availability resulted in leaf water potentials that were not very negative and none of the investigated species showed evidence of osmotic adjustment. The volume of water transpired by each of the species per unit land surface area was estimated from the relationship between abiotic factors and plant water loss. These relationships varied among species and had varying degrees of predictability as a result of differences in stomatal behaviour between the three species. The water requirements of A. populifolia and S. plumieri were adequately met by the water supplied by rainfall and the water stored in the dune sands. It was therefore not necessary to invoke the utilisation of ground water or the process of internal dew formation to supply sufficient water to meet the requirements. However, I. pes-caprae despite its lower transpiration rates and due to its higher biomass, lost greater volumes of water per unit dune surface area than either A. populifolia or S. plumieri. This resulted in periods of potential water limitation for I. pes-caprae. Incident light was the most important determinant of leaf photosynthetic CO₂ assimilation and transpiration, particularly as a linear relationship between incident PPFD and atmospheric vapour pressure deficit (VPD) could be demonstrated. Whole plant photosynthetic production by S. plumieri was shown to be light limited as a result of mutual shading despite high incident and reflected PPFD occurring in the foredune environment. The leaf hair-layer of A. populifolia was shown to be important in reducing transmitted UV and hence reducing photoinhibition but it also caused reduced transpiration rates because of the thicker boundary layer and thus increased leaf temperatures. The nutrient content of above-ground plant parts of the investigated species were typical of higher plants despite the low nutrient content measured for the dune soils. With the possible exception of nitrogen the nutrient demand created by above-ground production was adequately met by the supply of nutrients either from sand-water or from aquifer-water transpired by the plants. Differences in the volumes of water transpired, and hence the quantity of nutrients potentially taken up via the transpiration stream, resulted in interspecific differences in above-ground plant macronutrient content. The reallocation patterns of nutrients differed both between the various nutrients measured and interspecifically. Standing biomass and the density of plants per unit land area was low in comparison to that of other ecosystems and was different among investigated species. This may be important in maintaining the adequate supply of resources (water, nutrients and light). As a result of the interspecific differences in biomass when production was expressed per unit land surface area the resultant productivity was not dissimilar among species. Productivity was high when comparisons were made with species from other ecosystems. No single resource (water, nutrients or light) could be identified as the controlling factor in the foredune environment and a combination of both resource stress and environmental disturbance are likely to be involved. Physiology, production, growth and growth characteristics conveyed certain adaptive advantages to these species in respect to both resource stress and environmental disturbance. Interspecific differences in these adaptations can be used to offer explanations for the observed microhabitat preferences of the three investigated species. Furthermore features common to all three species offer some explanations as to why these species and not others are able to inhabit the foredunes.

M. Tech. (Department of Metallurgical Engineering, Faculty of Engineering and Technology), Vaal University of Technology. / The production of austenitic stainless-steel pump impellers in foundries present a huge challenge mainly due to its thin-walled blades, pouring temperature, presence of junctions and chemical composition. Two different alloys were used namely nodular cast iron and austenitic stainless-steel. Nodular cast iron was used as a comparison alloy due to its excellent flowability whereas austenitic stainless-steel was chosen due to its attractive corrosion and wear resistant properties. Austenitic stainless-steel alloy showed difficulties during casting because of its chemical composition and freezing range. Thin-walled sections are more susceptible to filling defects like misrun and cold-shut. This results in high scrap rate and high processing costs during high production of thin-walled components. High pouring temperature is considered one of the most effective methods to improve filling ability of thin-walled castings. However, there is a major drawback in using this method owing to the high occurrence of shrinkage defects and hot tearing especially at junctions. 1060 aluminium was used as a benchmark to evaluate the effect of wall thickness on the filling and feeding of thin-walled Al components with complex geometry during sand casting. The aim of this dissertation is therefore to optimize casting process of sand cast austenitic stainless-steel pump impeller. Numerical modelling and additive manufacturing were used to optimize the production of this product. The use of casting simulation software combined with three-dimensional (3D) mould printing technology has enabled optimisation of casting parameters to minimise the occurrence of casting defects. Casting parameters of five test samples of complex geometry and varying thicknesses (1.0 mm;1.5 mm;2 mm;2.5 mm and 3.0 mm) were optimised using MAGMAsoft® at a constant pouring temperature of 700 °C and 1060 Aluminium as an alloy. Simulation and casting results showed that complete filling was only possible at a wall thickness of 3 mm. The simulation results showed that as the wall thickness increased from 1 mm to 3 mm the filling ability increased by 67.5 % whereas experimental casting results showed that filling ability increase by 75 %. The combination of MAGMAsoft® simulation and 3D printed moulds proved to be effective tools in predicting filling and feeding of thin-walled aluminium components during sand casting. MAGMAsoft® casting software was used to simulate metal flow and predict the degree of filling at different pouring temperatures. Test samples were cast using 1060 Aluminium alloy at temperatures of 702 °C, 729 °C, 761 °C, 794 °C, 800 °C and 862 °C. Complete mould filling was predicted at 800 °C using the simulation model and 761°C during actual casting. At temperatures above 761°C tearing at the junction was quite pronounced. An optimal of 761°C pouring temperature was found to be appropriate pouring temperature when casting thin-walled aluminum components using sand casting. MAGMAsoft® casting software proved to be an effective tool in optimizing filling and feeding of thin-walled aluminium components during sand casting. Nodular cast iron pump impeller was optimized at 1500 °C using MAGMAsoft® and 3D mould printing technology. Design variables used were feeder radius (17 mm, 18 mm, 19 mm and 20 mm), feeder height (32 mm, 33 mm, 34 mm, 35 mm) and number of feeders of (3, 4 and 5). Simulation and casting results showed a completely-filled casting. The high fluidity of nodular cast iron promotes mould filling ability and prevent any form of misrun defect. Minimum shrinkage was noted at the junctions and top surface of the casting. A new design was proposed to eliminate shrinkage defects at the junctions of the nodular cast iron pump impeller. The design used a tapered circular runner bar with straight ingates. Optimization of nodular cast iron was now done at 1390 °C with the use of MAGMAsoft® and real casting was done 1385 °C. Simulation and casting were in correlation to each other since both showed completely-filled mould cavity with no misrun, cold-shut and shrinkage porosity defect. Simulation proved to be an effective tool in optimizing filling and solidification of nodular cast iron during sand casting. Austenitic stainless-steel pump impeller was optimized at 1500 °C using MAGMAsoft® and 3D mould printing technology. A high quality mould and core print were printed with the use of Voxeljet VX1000 at a minimum period of time. Design variables used were feeder radius (17 mm, 18 mm, 19 mm and 20 mm), feeder height (32 mm, 33 mm, 34 mm, 35 mm) and number of feeders of (3, 4 and 5). An increase in feeder size and the number of feeders greatly reduced hot spot and porosity of the casting but it also reduced the casting yield. The quality of the casting was found to be inversely proportional to the casting yield. Simulation showed a completely-filled casting with actual casting showing only 50 % filling ability. High viscosity of the molten metal and thin walled blades promote quick solidification which caused misrun defects. A new design was proposed to eliminate misrun defects of the first design. MAGMAsoft® was used to optimize this design at 1550 °C. The design used a tapered circular runner bar with tapered ingates. The actual casting showed improved filling ability from 50 % to 80 % while simulation showed completely-filled mould cavity (100 %). Major factors which contributed to low filling ability of austenitic stainless-steel pump impeller were chemistry, runner system and men. Numerical modelling and additive manufacturing did optimize filling and feeding of sand cast austenitic stainless-steel pump impeller.

Unique features of behavior of granular materials make constitutive modeling of these materials a challenge that has not yet been answered completely. Because volume changes are so important for the type of behavior exhibited by frictional materials, it is important to correctly incorporate them in constitutive models, both in terms of their rate of development and their magnitude. In this study a number of consolidated drained triaxial tests are performed to find those features of sand behavior that can be considered "material parameters" and can be used for constitutive modeling of granular soils. Special attention is given to those features of material behavior that are related to dilatancy. A number of published experimental data are also analyzed and useful trends of soil behavior are found.

The Lower Tertiary Balder Formation in the Beryl Embayment, North Sea, consists of sands interbedded with claystones and tuffs. The sands are massive and well sorted and can be up to 400 feet (122 m) thick, and are highly porous and permeable hydrocarbon reservoirs. The sands form large lobate and circular bodies of sands a few kilometres in diameter, with steep sides that are sometimes controlled by fault planes. The margins of the sands sometimes display thick sand wings extending up along fault planes. The sands display dewatering structures such as sills and dykes have a complex geometrical relationship with the surrounding sediments. Hydrothermal mineralisation is displayed as nodules, concretions and cementation of the sands within the Balder interval. The concept of seismic pumping, which postulates the rapid upward migration of deep fluids as the result of fault movement, was introduced to the literature some eighteen years ago, but fell into disrepute. However, it is argued here that re-shear of normal faults in the reverse direction can under certain critical physical conditions cause seismic pumping and can transport large quantities of deep seated fluids rapidly. This gives rise to the expulsion of fluid from depth into conventionally deposited massive sands of submarine fan environments, belonging to the Balder Formation, and thus in the fluidisation of the sediments. These sands have been intruded into the surrounding rocks and along fault planes forming a complex distribution of in situ and remobilised sands, thereby giving rise to the observed sand geometries and structures.

Oblique dunes have orientations that are intermediate between those of transverse and longitudinal dunes. The oblique dunes studied are reversing dunes which undergo no net annual migration when associated with normal meteorological patterns. From April 1980 through September 1980, the dunes migrated up to 65 feet (19.8 m) northwestward under the influence of prevailing onshore winds. High velocity northerly winds (November 1980 through February 1981), associated with the passage of winter frontal systems, caused the dunes to rapidly migrate 65 feet (19.8 m) southward. Volumes of sand transported by these strong winds were commonly reduced by accompanying rainfall. In October 1980 and March 1981 neither wind direction was dominant, and frequent changes in wind direction caused many of the dunes to become flattened. Hurricanes, which strike the area in late summer, have had no lasting effects on the dunes. Three major stratification types were observed in trenches and on etched surfaces. Translatent strata were deposited by wind ripples; grainfall deposits accumulated when saltating grains settled on leeward slopes of the dunes, and grainflow cross-strata were developed by avalanching on leeward slopes. Preservation of these stratification types occurred in zones of net deposition, predominantly leeward of the dune crests. Strata deposited during the summer wind regime dip northeast, whereas the winter strata dip in a southerly direction. The winter deposits are best preserved in the central cores of the dunes. This suggests that either the high velocity winds of the initial winter frontal systems destroy large volumes of the summer deposits, or that the dunes migrate southward, under the influence of dry northerly winds, during droughts. Oblique dune deposits should be difficult to discern in the rock record, because they may contain aspects of either transverse or longitudinal dunes. It is likely that some ancient oblique dunes have been mistakenly described as other dune types in the literature. / text

This study was motivated by the fact that many times intact specimens with a number of oversized particles are dynamically tested in the laboratory and the impact of the particles on the dynamic properties is unknown. The effects of oversized particles represented by gravel particles on the shear modulus (G) and material damping ratio (D) of a uniform sand were evaluated in the linear (γ ≤ 0.001%) and nonlinear (γ > 0.001%) ranges of shear strain with combined resonant column and torsional shear (RCTS) equipment. The sand used in this investigation is a uniform sand as a reference, well-characterized material on the dynamic properties. Sand-gravel specimens were constructed using the undercompaction method. A variety of rounded gravel particles was used in building the specimens. Dynamic tests on the sand-gravel specimens were performed, and the tests results are presented. Among the findings of this investigation are that, compared to uniform sand: (1) oversized gravel particles symmetrically located along the longitudinal axis in uniform sand generally decreased slightly the small-strain shear modulus (Gmax), (2) oversized gravel particles asymmetrically located away from the longitudinal axis of rotation resulted in slight increases in Gmax and the small-strain material damping ratio (Dmin), (3) the G – log γ relationships of sand-gravel specimens with asymmetrically located gravel particles are generally above those with gravel particles symmetrically located along the longitudinal axis, and (4) the G/Gmax – log γ relationships of all specimens were reasonably close for the nonlinear ranges covered in these tests (γ < 0.05 % and G/Gmax > 0.6). As long as the oversized particles were near the axis of rotation, the particles had little effect on the dynamic properties (Gmax, Dmin and G – log γ relationships) regardless of sizes and numbers of particles. However, once the oversized particles were located away from the axis of rotation and closer to the perimeter of the specimen, the oversized particles influenced the dynamic properties. Finally, the additions of oversized particles located both symmetrically and asymmetrically in the uniform sand specimens have little impact on the nonlinear dynamic properties (G/Gmax – log γ and D – log γ relationships) which compared well with uniform sand. / text

No description available.