The interface between the soil and structural material is the region where load is transferred for many geosystems, including deep foundations, soil nails, modified soil earth walls, and tunnel walls. Specifically, the frictional resistances generated by piles depend on a number of factors, such as the sand density and friction angle, effective stress acting on the pile surface, and the surface roughness of the pile. Certain types of piles shafts may benefit from resistances that are different in one direction of loading compared to the other, including reaction piles that generate lower installation resistance compared to the pullout resistance or offshore foundations subjected to tensile bias loading during a storm surge. Engineering the surface of such foundations to produce a directionally-dependent shaft resistance could result in a reduction of the required installation depth, and therefore reducing material and installation costs. A pile shaft that results in directionally-dependent resistances may be designed by structuring the shaft surface with asymmetric asperities similar to a ratchet. In order to narrow the parametric space of possible solutions, a bioinspired approach was employed in this work. Biogeotechnics has gained traction as a field in recent years due to its utility in providing efficient solutions to engineering problems based on designs already existing in nature. A number of morphologies exhibit directional-dependent resistances, including the ventral scales of snakes, which were considered in these studies. Particularly, cranial shearing is induced when the scales are displaced against the soil, resulting in mobilization of larger frictional resistances. In contrast, caudal shearing is induced when the scales are displaced with the soil, resulting in smaller frictional resistances. A number of geometric profiles based on the ventral scales of snakes were used to study pile shaft interface behavior, both as planar surfaces in laboratory interface shear testing and as piles with custom-machined surfaces. Constant normal stiffness (CNS) laboratory interface shear testing on sand was performed on reference rough, smooth, and snakeskin-inspired surfaces to study the interface shear at the unit resistance level under both monotonic and cyclic loading, while centrifuge pile load testing was performed to assess the analogous field-scale behavior in terms of installation and pullout resistances as well as stability under cyclic loading. The CNS laboratory tests showed that the snakeskin-inspired surfaces mobilized directional-dependent shear resistances under monotonic loading, which generally agreed with results from the pile shaft resistances measuring during installation and pullout. The laboratory tests also showed that the failure envelopes associated with the snakeskin-inspired surfaces were nonlinear compared to reference rough and smooth surfaces. Potential mechanisms underlying this trend were investigated using particle image velocimetry and it was found that the trend is correlated with the shear strains at the interface at different normal stress conditions. Pile load tests were conducted on internally instrumented piles with reference rough, smooth, and snakeskin-inspired pile shaft geometries to assess the evolution of load transfer with depth. The results indicate that piles displaced in the cranial direction gradually shed load compared to piles displaced in the caudal direction due to the greater magnitudes of mobilized skin friction of the former. In addition, cranially displaced piles require more displacement to reach the maximum resistance compared to caudally displaced piles. When the piles were subjected to cyclic loading with a tensile bias, it was found that the cranially pulled piles failed in fewer cycles than the caudally pulled piles when normalized by the total shaft resistance but could resist greater absolute loads. The results highlight the effect of both cranial and caudal directions on cyclic stability of pile shafts. Cyclic interface shear tests were conducted using CNS conditions to better understand the mechanics governing the pile shaft behavior. The results indicate that under symmetric loading conditions, pile shaft elements fail in the caudal direction in a brittle manner. However, when a mean load bias is introduced in the cranial direction, the behavior changes and the interfaces fail by a progressive accumulation of displacements in the caudal direction. The results also provide insight to the effects of initial normal stress, loading amplitude, and boundary stiffness on the number of cycles to failure for both reference rough, smooth, and snakeskin-inspired surfaces. These studies show the applicability of snakeskin-inspired geometries in piling applications. The agreement between laboratory and centrifuge pile tests indicate that trends observed in laboratory tests may inform field scale behavior. The use of the snakeskin-inspired surfaces readily exhibits directional-dependent load transfer behaviors which may be used to reduce or increase resistances in compression or pullout depending on the design requirements.

Soil penetration activities, such as site investigation, pile driving and tunneling, are ubiquitous and fundamentally important in civil engineering. These penetration activities are energy-intensive and responsible for considerable environmental disruption due to the usage of large equipment required to provide reaction force. Bio-inspired geotechnics has received significant attention in recent years due to the growing need for making construction processes more sustainable. This research investigates the anchor-tip strategy inspired by earth and marine worms, razor clams, and caecilians and the circumnutation strategy inspired by plant roots. Throughout this dissertation, it is shown that using the bio-inspired strategies can facilitate soil penetration by reducing the mobilized penetration resistance and generating anchorage reaction forces.Discrete element modeling (DEM) is used to simulate the bio-inspired soil penetration process of an in-situ testing probe in granular soils. Different soil conditions are examined in the DEM simulations, including soil density and magnitude of overburden stress. Simulations were performed on specimens confined to constant stress levels using servo-controlled algorithms to model deep penetration conditions. Simulations were also performed on unconfined soil specimens under gravity to model shallow soil conditions. The simulated anchor-tip strategy consists of radially expanding of a probe section or sections (i.e. anchor(s)) and subsequently displacement of the probe tip to deeper locations. The global and meso-scale responses of the probe-soil system are analyzed to shed light on the working mechanisms of this strategy. Specifically, the expanded anchor serves as a integrated anchorage system to provide the reaction force needed for penetration. The anchor expansion leads to reduction in tip resistance by altering stress states around the tip. The effects of a number of aspects of the anchor and tip geometry, as well as of soil depth, are explored. The effects of soil density on the anchor-tip strategy are also highlighted. Due to the complexity of anchor-tip strategy, three simulation strategies employed to control the probe’s motions: displacement-controlled algorithm, velocity controlled algorithm with force limits and force-controlled algorithm. Among these three strategies, the first one is the most simplified one while the last one best approximated the motion used by the model organisms. The simulated circumnutation inspired motion (CIM) consists of helical movements of the probe tip accompanied by downward penetration of the entire probe. The probe forces, torque, mechanical work and particle contact orientations are analyzed and discussed. The results indicate that CIM leads to a decrease of penetration resistance by altering the contact orientations near the tip from the vertical to the horizontal direction. However, this reduction in penetration resistance comes at a cost of increased torque and in most conditions an increase in rotational work. A comparison of the CIM and rotational penetration (RP) strategies shows that the CIM mobilizes smaller penetration forces and requires less total mechanical work to penetrate the same distance as the RP. The effects of the CIM velocities and probe geometry are also examined. The understanding of the probe-soil interactions during the bio-inspired penetration processes lays the foundation for the development of innovative soil penetration tools and techniques to increase the efficiency of construction activities. For example, studies on bio-inspired probes can guide the design of future lightweight penetration devices. Such probes could reduce the energy consumption during transport and the challenges associated with limited accessibility in certain sites, such as congested urban areas and outer space bodies. These studies show that by learning from nature, more efficient solutions can be developed for geotechnical engineering applications.

The application of bio-inspiration in geotechnical engineering can help develop more efficient and effective solutions. Biological processes, such as the growth of tree roots, have evolved to accommodate limited resources for growth. Recent work provides evidence that root systems can have a significantly higher axial load capacity per material unit mass or volume than conventional foundation systems such as micropiles. For the work presented in this thesis, two small scale root-inspired models (6-leg and 3-leg), a pile, and a plate anchor were 3D printed and tested under combined loading conditions. The tests were performed using a six-axis robotic arm at 1g in subrounded, loose sand. The results of these tests were compared to assess the performance of the tree root-inspired models relative to more standard shapes. In both vertical and horizontal pullout, the piles had the lowest absolute pullout resistance and the plate anchors the highest. The results showed the mobilization of peak capacities at smaller displacements as the pullout angle became more vertical. After normalizing the results by volume, the tests indicate a greater material efficiency in the tree root model with 6 legs relative to the plate, the pile, and the 3 leg models for most directions of loading. The pile gradually increased peak capacity as the pullout angle approached horizontal, nearly matching in material efficiency with the 6-leg model for the horizontal load test. The results suggest that tree-root inspired shapes may provide anchorage with lower material use than traditional practice. In addition, the results of this study show that 1g tests using a robotic arm be used to explore the behavior of soil-structure interaction problems.

Inherent particle properties such as size, shape, gradation, surface roughness and constituent material (i.e. mineralogy) control the particle-scale contact response of particles which in turn determines the global mechanical behavior of coarse-grained soils observed in both laboratory tests and full-scale geosystems. This dissertation presents a series of studies as part of two separate projects aimed to characterize the effects of particle shape and gradation on the mechanical behavior of coarse-grained soils.

Systematic investigation of the effects of an individual particle property on the mechanical behavior of granular soils is a pervasive challenge in experimental studies with naturally occurring soils because it requires careful control over the remaining particle properties. Due to this challenge, conflicting interpretations of the effect of different particle properties on the behavior of soils exist in literature. By taking advantage of state-of-the-art additive manufacturing (i.e. 3D printing) technology, artificial sand analog particles can be manufactured with independent control over different particle properties such as size, shape and gradation. The first part of this dissertation investigates the feasibility of 3D printing technology to model the mechanical behavior of coarse-grained soils. The results of this study show that 3D printing technology can be used successfully to create artificial sand analogs with different sizes and shapes using either X-ray CT scans of natural sands or synthetic shape generating algorithms based on spherical harmonics. Although the 3D printed analogs exhibited greater compressibility compared to that of natural sands, the shear wave transmission behavior of 3D printed sands, measured using piezoelectric bender elements, exhibited dependencies on mean effective stress, void ratio and particle shape that are quantitatively similar to those previously reported for natural sands. By analyzing the shear wave transmission data, equations to predict the shear wave velocity taking into account the particle shape and void ratio were developed. The triaxial compression behavior of the 3D printed sands was investigated in both drained and undrained conditions, which exhibited a stress-dilatancy response typical of natural sands, and the interpretation of their shear response can be captured within the critical state soil mechanics framework. The results of tests on 3D printed sands show that changes in particle shape produce changes in friction angles and critical state parameters that are similar to those observed in natural sands. However, the greater compressibility of the 3D printed material and the smaller inter-particle friction coefficient should be considered in the interpretation of results. The use of methods based on poorly-graded sand data to characterize the strength and stress-dilatancy behavior of widely-graded soils is a common practice in geotechnical engineering; however, many naturally-occurring soils encountered in the field are widely-graded as evidenced by case studies. The second part of this dissertation examines the effect of gradation and particle size on the strength and stress-dilatancy behavior of widely-graded coarse-grained soils. A well-graded natural sand sourced from the Cape May Formation near Mauricetown, New Jersey was selectively sieved to produce different sands with similar particle shape parameters but with different gradation and median particle size. The results of a series of isotropically consolidated drained and undrained triaxial tests exhibited dependency on the peak strength, dilatancy and critical state parameters of the gradation and median particle size. The analysis of the results also showed that capturing the effects of gradation and particle size depends on the definition of soil state, where the state parameter provides more robust trends than the relative density. The results indicate that for any given state parameter, more widely-graded soils exhibit greater peak friction angles, maximum dilation angles, and differences between the peak and critical state friction angles.

The research efforts presented in this dissertation show that the individual effects of particle shape, size and gradation on the soil behavior can be investigated using new experimental techniques (i.e. creating artificial sand particles using 3D printing) or selective sieving of naturally occurring soils. This can bring benefits in improved understanding of soil behavior aiding increased efficiency and robustness of geotechnical site characterization and design methodologies and in the advancement of constitutive models.

The failure of impounded fly ash at the Kingston Fossil Plant in 2008 led to severe economic and environmental consequences from the flow of ash for almost one kilometer downstream. With over 700 active surface impoundments and federal disposal regulations enacted in 2015, the coal-power industry is required to properly address the risk and possible consequences associated with impounded fly ash. The Electric Power Research Institute and the University of California Davis, Center for Geotechnical Modeling collaborated to advance the understanding of possible flow failures of impounded ash, including conditions that lead to its triggering, methods to identify its potential in the field, and the potential impacts of dewatering. A centrifuge testing program began in 2018 to explore the influence of initial density and dewatering of impounded ash, with a focus on its runout behavior after a model impoundment is subjected to a loss of lateral confinement. This study presents the results from the one centrifuge test performed at 60 g and three 1 g tests conducted in 2021, with comparisons made to four previous 60 g tests. Fly ash was deposited into a novel centrifuge container, with a pair of gates opened in flight to induce an idealized containment structure failure. The strength and stiffness of the ash were measured prior to the failure using a miniature CPT penetrometer and bender elements for the 60 g test and a miniature T-bar penetrometer for the 1 g tests. Pore pressures were measured throughout the entire 60 g test using tensiometers, and water contents were measured using moisture probes for all tests during specimen preparation and testing. Various cameras were used to track the runout after the opening of the gates. Failure mechanisms of the ash ranged from slope instability to a drastic flow failure in the 60 g tests, and flow failures to a slumping-like failure in the 1 g tests Although the runout results between the 60 g and the 1 g tests agree in the effect of ash density, the failure mechanisms were not found to be comparable due to differences in confining stress between the two test types. Ash strength and stiffness and water table height were found to impact the runout length in all tests. The results from this testing program and comparison to the previous centrifuge tests demonstrate that dewatering can improve the stability of fly ash impoundments and that looser deposits are more prone to flow failures associated with large runout distances.

Numerous geotechnical applications such as pile installation or CPT exploration require large rigs to generate reaction forces for the penetration of geotechnical elements. These rigs can increase financial and environmental impacts and pose accessibility challenges to engineering projects. Bio-inspiration can be used to identify geometries of penetrometers to reduce the penetration resistances, which may enable the use smaller rigs that would reduce the related economic and environmental impacts of geotechnical activities. This research aims to identify the attributes of organisms that make them efficient burrowers and evaluate the application of these attributes for geotechnical engineering activities. Particularly, this work focuses on the attributes of penetrometer apex angle and geometric asymmetry. Evaluation of the bio-inspired penetrometer, in terms of the generated tip resistance, is accomplished by performing Discrete Element Method (DEM) simulations where each geometry is penetrated into a specimen contained in a calibration chamber. In this research, both shallow and deep penetration conditions are considered. Shallow penetration conditions are defined as penetration in an unconfined specimen at normalized depths ratios (Z/D, depth to probe diameter) smaller than 7, while deep penetration conditions are defined as penetration in a confined specimen subjected to a vertical stress of 100 kPa and a horizontal stress of 50 kPa. Results show that an apex angle close to that of a honeybee stinger, 30°, minimizes the tip penetration resistance for shallow penetration. Results additionally show that an apex angle of 15° reduce penetration resistance for deep penetration conditions. However, this reduction in penetration resistances at deep conditions was smaller than that achieved in shallow conditions. With DEM, it is possible to monitor the forces and position of each individual particle, thus allowing for close examination of the soil failure mechanisms generated by each tested probe geometry. The results show that probes with small apex angles displace particles horizontally and create increases in horizontal stress at locations near the probe tip. In contrast, probes with large apex angles displace particles vertically down, creating an increase in vertical stresses below the probe tip. The asymmetric probes simulations showed no reduction of penetration resistance compared to their symmetric counterparts. The asymmetric tip may also contribute to an imbalance in the forces acting on the probe, which may cause the probe to not penetrate the substrate vertically. The mechanisms explored using DEM can also help develop an understanding for future improvements of probe geometry to achieve more efficient penetration during in-situ tests and construction activities.

The inherent particulate nature of granular soils, such as sands and gravels, plays an important role in their engineering behavior. This dissertation aims to advance the fundamental understanding of the effect of fabric- and stress-induced anisotropy, and gradation on the micro- and macro-scale behavior of coarse-grained granular soils.The first portion of this dissertation addresses the direction-dependent characteristics exhibited by soils reflected in the anisotropy of their responses. Studies have shown that both the depositional processes and particle arrangements (i.e., fabric-induced anisotropy), and the stress state and history (i.e., stress-induced anisotropy) impact the anisotropic behaviors observed at the macroscopic level. Quantifying these anisotropies has been challenging, necessitating specialized geotechnical testing and imaging equipment. To overcome these challenges, a novel experimental testing setup is introduced, designed to measure shear wave velocities (VS) along different orientations and polarization planes using piezoelectric bender elements (BEs) to obtain angular distributions of VS. Subsequently, two investigations on shear wave propagation are presented, using the developed setup and Discrete Element Method (DEM) simulations, to explore the effects of fabric- and stress-induced anisotropy on the VS anisotropy. The experimental tests were performed on glass beads and angular natural sands, while the DEM simulations used spherical and rod-like clumped particles. These specimens were subjected to isotropic and one-dimensional (1D) compression. The results reveal that the angular distributions of VS and measurements obtained along different polarization planes (i.e. VS,HH, VS,HV, and VS,VH) can discern the effects of fabric and stress anisotropy. The observed trends indicate a relationship between the angular distributions of VS and of the alignment of particles and interparticle contact forces. A framework is presented based on the VS measurements along various orientations and polarization planes which is validated using the presented results. When presented in terms of the ratio of VS measurements along different orientations and polarization planes, namely the VS,HV/VS,VH and VS,HV/VS,HH ratios, and of the newly introduced Anisotropy parameter (Ae), this framework facilitates the evaluation of the stress- and fabric-induced anisotropy in soil specimens. The results also highlight the challenges in discerning the effects of stress and fabric anisotropy when both simultaneously influence the soil response. Geosystems built on coarse-grained soils with broader gradations are typically designed and analyzed using methodologies developed for poorly-graded soils without explicit consideration of the effects of gradation, potentially leading to uncertainty in performance predictions. In the second portion of this dissertation, the effects of changes in the gradation on various aspects of monotonic and cyclic response of coarse-grained soil behavior are investigated using DEM simulations. The simulations include monotonic isotropically-consolidated drained and undrained triaxial tests, and cyclic undrained direct simple shear tests conducted on specimens with coefficients of uniformity (CU) between 1.9 and 6.4 composed of non-spherical particles. The triaxial simulation results indicate that an increase in CU leads to increases in peak shear strength, dilative volume change, rate of dilation, negative pore pressure generation, and rate of pore pressure generation. These findings are compared with established frameworks to highlight the differences in response resulting from variations in CU. Particle-level measurements from the monotonic simulations highlight the influence of gradation on both the packing characteristics and the transmission of contact forces within the soil assembly. In particular, for the broadly graded specimens, the coarsest particles exhibit a disproportionately higher number of connections and carry significantly greater contact forces compared to the coarsest particles in poorly graded soils. The coarsest particles for the broadly graded specimen are connected to a disproportionally higher number of particles and carry disproportionally higher contact forces as compared to coarsest particles in poorly graded soils. The enhanced interlocking of the coarser particles results in greater dilation during shearing, leads to higher peak shear strengths for the more broadly graded specimens. Additionally, the particles smaller than D10 are inactive in contact force transmission, while the percentage of particles active in contact force transmission increases with an increasing CU. During cyclic shearing, specimens with broader gradations yield lower liquefaction-triggering resistance than poorly graded specimens at similar relative densities. Conversely, the opposite trends emerge when compared at similar initial state parameters. Post-liquefaction, specimens with broader gradation accumulate shear strains at a smaller rate. A comparison is presented, examining the interpretation of grading-dependent behavior by choosing relative density or initial state parameters as the state definition for both monotonic and cyclic response, highlighting the efficacy of initial state parameter in capturing the systematic differences in the response because of changes in gradation. The improved interlocking in more broadly graded specimens results in a lower percentage of sliding contacts for both strong and weak force-carrying contacts at the initiation of liquefaction and in subsequent cycles, which is linked to the slower rate of post-liquefaction strain accumulation in well-graded specimens. The combination of macro and micro observations, from the research efforts presented in this dissertation, highlight the influence of fabric- and stress-induced anisotropy and gradation on soil behavior through a combination of novel experimental testing and DEM simulations, and contribute to the advancements in the geotechnical site characterization, design methodologies, numerical simulation techniques, and constitutive modeling of granular soils.

We present here a first comprehensive database on the diversity of proseriate flatworms (Platyhelminthes: Rhabditophora: Proseriata) on Western Mediterranean microtidal, wave dominated beaches. We sampled 116 stations in two years, through Spain (22 beaches, including Balearic Islands), France (25 beaches, including Corsica), Italy (63 beaches, including Sardinia, Sicily, and Lampedusa), and Tunisia (6 beaches). In each beach, we sampled at three depths, corresponding to the swash, shoal, and subtidal zones. For each sample, we obtained environmental data. The research yielded a total of 152 species, of which 93 were new to science. For each of the species found, we coded and described 16 functional traits. We discuss the functional meaning of the selected traits, as well as on diversity patterns and emerging biogeographic signals across the investigated regions. We particularly focused on the most widespread and dominant species in our dataset, concentrating on their putative adaptations to high energy environments; as well as the high number (58) of the species only found once. Finally, we discussed the coverage of our sampling by estimating the diversity at each investigated region and comparing it to the actual diversity. All information provided is available through the Global Biodiversity Information Facility (GBIF) and the Open Science Framework (OSF) following the Darwin Core Standard.

Knee osteoarthritis is a degenerative musculoskeletal disorder marked by gradual cartilage breakdown, and involving all tissues of the joint. It is one of the most common worldwide causes of chronic disability in older populations, with the prevalence expected to increase. There is currently no available treatment to reverse the degenerative damage characteristic in osteoarthritis, with the only option available for end stages of the disease being a partial or total knee replacement. Furthermore, the clinical standard for osteoarthritis diagnosis is a radiographic score which reflects advanced pathological stages, often with irreversible damage. The lack of therapies has generated a need for osteoarthritis imaging biomarkers capable of detecting and monitoring the progression of the disease. This dissertation aims to bridge this gap by defining a novel spherical encoding representation for known quantitative imaging biomarkers for osteoarthritis. In this work, we leverage the superior cartilage sensitivity of MRI, a large retrospective labeled imaging dataset, and the superior feature-learning ability of convolutional neural networks to define novel OA imaging biomarkers based on spherical maps. Large-scale quantitative analysis using convolutional neural networks uncovered new associations between bone shape, cartilage thickness, and cartilage T2 relaxation time values and OA symptoms.