Idag blir det allt vanligare med träbyggnader högre än två våningar. Detta tillsammans med en modern arkitektur som ger stora öppna planlösningar ställer höga krav på bjälklagen i träbyggnader. Problematiken med långa spännvidder för bjälklag i trä är att klara kraven för svikt och nedböjning. Dessa krav måste uppfyllas för att säkerställa funktioner hos andra byggdelar och för att människor inte ska uppleva att golvet sviktar eller vibrerar på ett obehagligt sätt. Ett träbjälklag bestående av limmade balkar av sidobräder har utvecklats. Bjälklaget är utformat av balkar med I-tvärsnitt i primärriktningen och rektangulära balktvärsnitt i sekundärriktningen. Examensarbetet omfattar laborativa provningar och beräkningar dels för att bestämma en böjelasticitetsmodul för varje enskild limmad balk och dels för att bestämma styvhetsegenskaperna för bjälklaget. De limmade träbalkarna ingår i ett forskningsprojekt vid Växjö universitet finansierat av KK-stiftelsen, som syftar till att undersöka möjligheterna att tillverka en konkurrenskraftig produkt genom att i grönt tillstånd (otorkat) limma ihop bräder från stockens yttre delar till balkar. Balkarna levererades limmade och hyvlade till universitet där en böjelasticitetsmodul först bestämdes för varje enskild balk. Därefter monterades balkarna ihop till ett fullskaligt bjälklag som provades med olika försöksuppställningar/lastfall varvid deformationen mättes upp. Dessa deformationer blir underlag för att bestämma bjälklagets styvhet. Böjstyvheten i primärriktningen uppgår till 17,55 x 1012 Nmm2/m enligt beräkningar baserade på laborativa resultat. Böjstyvheten i sekundärriktningen uppgår till 4,5 % av primärriktningens böjstyvhet, dvs. 0,79 x 1012 Nmm2/m. Sammanfattningsvis kan man säga att böjstyvheten är hög i båda riktningar i jämförelse med vanliga träbjälklag. / In Sweden it becomes more and more common with wood buildings higher than two floors. This along with a modern architecture that gives big open plan solutions sets high requirements on the floor systems in wood buildings. The complexes of problems with long spans for floor systems in wood are to match the requirements for elasticity and deformation. These requirements must be met in order to ensure functions of other construction components, and not be unpleasant for people to walk on with respect to vibrations. A wooden floor system consisting of green glued side wood sections has been developed. The floor system is made with I-profiled beams in the primary direction and rectangular cross-sections in the secondary direction. This diploma work is based on that through elaborative testing and numeric calculations to decide the stiffness properties for each individual green glued side wood section and for the floor system. The glued side wood sections are included in a project at Växjö University, which is financed by the KK-foundation. The sections were delivered glued and planed to the university where the stiffness properties were first determinded for each individual section. Then, the sections were assembled to one fully sized floor system that was exposed to different experiments as the deformation was measured. These deformations were later used in order to decide the stiffness of the floor system. The stiffness in the primary direction was prescribed to 17,55 x 1012 Nmm2/m after calculations using results from the tests. The stiffness in the secondary direction amounts to 4,5% of the primary directions stiffness, i e. 0,79 x 1012 Nmm2/m. To sum up, one can say that the stiffness is high in both directions compared to regular wooden floor systems.

I samband med att intresset för att bygga högre trähus har ökat så krävs nya lösningar för att t.ex. kunna möta efterfrågan på stora öppna ytor och långa spännvidder. Träbjälklag med lång spännvidd har dock oftast svårigheter med att klara kraven på svikt och vibrationer. Ett nyutvecklat förslag på träbjälklag som förmodas klara dessa krav bättre än traditionella träbjälklag har varit utgångspunkten för detta examensarbete där syftet har varit att undersöka bjälklagets styvhet. Detta gjordes laborativt genom att bygga och testa en prototyp av det föreslagna bjälklaget och genom att en numerisk modell baserad på finita element metoden togs fram och användes för att studera hur olika parametrar påverkar bjälklagets styvhetsegenskaper. Bjälklagets design bygger på fackverksprincipen i primärriktningen och på balkverkan i sekundärriktningen. De ingående komponenterna har kommit prefabricerade till Växjö universitet där de har monterats ihop till ett fullskaligt bjälklagselement. Elementet har sedan utsatts för ett antal belastningsfall där nedböjningarna uppmätts vilka sedan givit underlag för att få värden på bjälklagets effektiva styvhetsegenskaper. Både de laborativa och de simulerade resultaten visar på en hög böjstyvhet i primärriktningen d.v.s. 18,9•106 Nm2/m [EI/b] respektive 18,6•106 Nm2/m [EI/b]. Även böjstyvheten i sekundärriktningen är hög d.v.s. motsvarar 21,2 % respektive 17,1 % av styvheten i primärriktningen. I beräkningsmodellen har det dessutom undersökts hur ett övre lager av spånskivor inverkar på bjälklagets styvhet. / The interest for building higher and larger wooden houses has increased in Sweden during the last decade resulting in higher requirements on the technical performance of such structures in order to met demands on large open surfaces and large spans of floors. Wooden floor systems with large spans often have difficulties, however, to meet the vibration requirements. A novel floor system, likely to handle the vibration requirements better than traditional wooden floor systems, is the basis for this master thesis. The purpose is to examine the stiffness of the floor by building and testing a prototype and by producing a numerical model based on the finite element method. In the longitudinal, main load-bearing direction the floor system works as a truss with flanges of longitudinal oriented timber members and web diagonals of transversely oriented members. In the transverse direction the web diagonals work as beams. The components were prefabricated elsewhere and assembled at Växjö University into a prototype. The prototype was then exposed to a number of different load cases. Deflections were measured and stiffness properties of the floor were derived. In addition to the experimental analysis the numerical model was used to calculate deflections when subjected to different load cases and for evaluating the principal stiffness properties of the floor. Both the experimental and the calculated results using the numerical model show high bending stiffness in the longitudinal direction, EI/b = 18,9•106 Nm2/m and 18,6•106 Nm2/m respectively. Also the bending stiffness in the transversal direction is high and equivalent to 21,2 % or 17,1 % (testing and simulation respectively) of the bending stiffness in the longitudinal direction. Using numerical analysis, also the effect on the stiffness of adding an upper layer of a 22 mm particleboard was examined.

Fibre-to-board is a simulation model developed at Stora Enso Research Centre Karlstad. Within this model isotropic hand sheet properties are used as input data for prediction of the final multi ply board properties. In order to improve and verify the calculations from simulations in Fibre-to-board so that these will correspond better with the results from the measurements on the paper/board machine, it was requested at RCK to investigate the possibility to optimize the input data to the model. Standardized hand sheet forming always results in sheets with properties far away from those produced on a machine. Therefore the aim with this Master thesis was to modify the laboratory procedure to receive hand sheets with properties closer to machine sheets. To achieve this, it was investigated how different parameters affect the sheet properties and if the hand sheet making process could be improved. When freely dried sheets were investigated it was found that sheets pressed with a wire clothing between the blotting paper and the hand sheet were less cockled than sheets pressed against only blotting papers. These sheets also tend to have a higher density. The cockling i.e. as a result from shrinkage was also reduced when the sheets were dried between slightly weighted wire clothing. Neither wire clothing nor orientated blotting papers during pressing eliminate the influence of anisotropic blotters on the shrinkage for isotropic hand sheets. It was also examined how the fine material influences sheet properties. The results showed that an increase in fines content result in higher shrinkage, higher density, increased TSI, more cockling and decreased air permeability. Different pressing loads and an increased density did not have much influence on the shrinkage. The density for freely dried sheets increased with higher load, but the results did not reach machine sheet densities, when the laboratory platen press was used. It might be difficult to receive freely dried hand sheets with higher densities. This is because fibres in freely dried sheets tend to relax after pressing, which will influence the density. Another press than the platen press used in these studies might compensate this matter. An increased pressing load resulted in less cockled sheets. The basis weight did not seem to have that large affect on the shrinkage when using machine chest furnish, therefore the basis weight on hand sheets used as input data to the simulation model Fibre-to-board might not be that important. It was studied how different plies and SW/CTMP pulp in a mixture affect the shrinkage. The results showed that the shrinkage increased with a higher SW content. It was also found that there is a linear relation between the total shrinkage of a SW/CTMP pulp mixture and the shrinkage for each individual pulp. In order to verify the Fibre-to-board model a simulation finally was performed. Furnishes and CD profiles of board were collected from a particular board machine within the Stora Enso Group. Properties from hand sheets made of furnishes were used as input data and the machine CD profiles were used as references. The CD TSI value corresponded with the value received from measurements on the machine board, but the MD TSI value did not. The shrinkage calculated on machine sheets did not coincide with the shrinkage from the simulation in Fibre-to-board. There are insecurities in the results from shrinkage measurement on the board CD profile due to the lack of width measurement during the process, which complicates the validation of the Fibre-to-board model. / Fibre to board är en simulerings modell framtagen vid Stora Enso Research Centre Karlstad. Modellen används för att prediktera krympning och styrkeegenskaper hos en bestämd kartongbana. Indata till modellen hämtas ifrån isotropa laboratorieark. För att förbättra och verifiera erhållna resultat från simuleringsmodellen så att de korresponderar bättre med värden från kartongmaskinen fanns det ett önskemål från RCK om att undersöka möjligheten att optimera indata till modellen. Laboratoriearktillverkning enligt standard resulterar alltid i ark med egenskaper som ligger långt från maskin arkens. Därför är syftet med detta examensarbete att modifiera arktillverkningsmetoden så att laboratorieark med egenskaper närmare de för maskinark kan erhållas. För att lyckas med detta undersöktes det hur olika parametrar påverkar pappersegenskaperna och om tillverkningsmetoden kunde förbättras. Vid undersökning av fritorkade ark upptäcktes att ark som pressats med viraduk mellan läskark och laboratorieark blev mindre buckliga än ark som pressats med enbart läskark. Dessa ark hade också en något högre densitet. Buckligheten som är en följd av krympningen reducerades också när arken torkades mellan viraduk under lätt belastning. Läskarkens inverkan på de isotropa arken kunde inte elimineras genom att använda viraduk vid pressning, inte heller genom att växla läskarken så att deras MD riktning orienterades olika. Även finmaterialets inverkan på pappersegenskaperna undersöktes. Resultaten visade att ett ökat finmaterial innehåll ger ökad krympning, högre densitet, ökat dragstyvhetsindex, buckligare ark och en minskad luft permeabilitet. Det visade sig att olika presstryck ger arken en högre densitet men krympningen påverkades inte märkbart. Densitet i samma nivå som på maskinark kunde däremot inte erhållas med laboratorieplanpress. Detta kan bero på att fibrerna i fritorkade ark relaxerar efter pressning, vilket ger en lägre densitet. För att kunna få högre densitet kanske en annan press än den planpress som användes i dessa studier kan införas. Det kunde även konstateras att ett ökat presstryck ger mindre buckliga ark. I dessa studier, där ark tillverkades av färdiga skiktblandningar från maskinkar, hade inte ytvikten på arken någon större inverkan på krympningen. Detta tyder på att ytvikten på arken som används som indata inte har så stor inverkan vid simulering i modellen Fibre-to-board. Det undersöktes även hur skikten i ett två-skikts ark och en blandning av LF/CTMP massa påverkar krympningen. Resultaten visade att krympningen ökar med en högre andel LF och att det finns ett linjärt samband mellan den totala krympningen för en blandning av LF/CTMP massa och krympningen för de enskilda massorna. För att kunna verifiera beräkningsmodellen Fibre-to-board utfördes slutligen en simulering. Skiktblandningar och tvärsprofiler från kartong togs ut från en specifik pappers maskin inom Stora Enso koncernen. Egenskaperna på laboratorieark gjorda av skiktblandningarna användes som indata till simuleringsprogrammet och kartongprofilerna från pappersmaskinen användes som referens. Dragstyvhetsindex i CD stämde bra överens med de mätningar som gjordes på maskinarken, men dragstyvhetsindex i MD skiljde sig. Krympningen som beräknades på maskinarken överensstämde inte med det simulerade resultatet. Det förekommer en osäkerhet i krympmätningarna som gjordes på kartong profilerna, då det idag inte förekommer någon mätutrustning på pappersmaskinen, som bestämmer bredden mellan press- och torkpartiet. Detta komplicerar valideringen av Fibre-to-board modellen.

There are several reasons to research the effects of axial twist exposures and the resulting loading on the spine. The lack of consensus from the limited work that has previously examined the role of axial twist moments and motions in the development of spine injuries or generation of low back pain is the primary reason. From recently published works, axial twist moments appear to represent an increased risk for injury development when it acts in concert with loading about other physiological axes (i.e. flexion, extension, and compression). However, there is a large body of epidemiologic data identifying axial twist moments and/or motion as risk factors for low back disorders and pain, demonstrating the need for this series of investigations. It is likely that these combined exposures increase risk through altering the spine’s load distribution (passive resistance) by modifying the mechanics, but this deduction and related causal mechanism need to be researched. The global objective of this research was focused on determining whether there is evidence to support altered load distribution in the spine, specifically between the intervertebral disc and facets, in response to applied axial twist moments (when added in combination with one and two axes of additional loading). Also included was whether these modes of loading can modify spine mechanics and contribute and/or alter the development of damage and pain. This objective was addressed through one in-vivo (Drake and Callaghan, 2008a– Chapter #2) and three in-vitro (Drake et al., 2008– Chapter #4; Drake and Callaghan, 2008b– Chapter #5; Drake and Callaghan, 2008c– Chapter #6) studies that: (1) Quantified the amount of passive twist motion in the lumbar spine when coupled with various flexion-extension postures; (2) Documented the effects of flexion-extension postures and loading history on the distance between the facet articular surfaces; (3) Evaluated the result of axial twist rotation rates on acute failure of the spine in a neutral flexion posture; and (4) Explored whether repetitive combined loading has the ability to cause enough deformation to the spine to generate pain. Through the combination of findings previously reported in the literature and the outcomes of Drake and Callaghan (2008a– Chapter #2) and Drake et al. (2008– Chapter #4), a postural mediated mechanism was hypothesized to be responsible for governing the load distribution between the facet joints and other structures of the spine (i.e. disc, ligaments). Increased flexed postures were found to decrease the rotational stiffness by resulting in larger twist angles for the same applied twist moment in-vivo relative to a neutral flexion posture (Drake and Callaghan, 2008a– Chapter #2). This suggested there might be an increased load on the disc due to a change in facet coupling in these combined postures. Similarly, increased angles were observed in flexed and twisted postures for in-vitro specimens relative to a neutral flexion posture. These observed differences were found to correspond with altered facet joint mechanics. Specifically that flexed twisted postures increased the inter-facet spacing relative to the initial state of facet articulation (Drake et al., 2008– Chapter #4). These finding supported the postulated postural mechanism. Therefore, in a neutral posture the facet joints likely resisted the majority of any applied twist moment based on the limited range of motion and higher axial rotational stiffness responses observed. It was suspected that the changes in mechanics would likely cause a change in the load distribution however the magnitude of change in load distribution remains to be quantified. Further support for this postulated postural mechanism comes from the mode of failure for specimens that were exposed to 10,000 cycles of 5° axial twist rotation while in a static flexed posture (Drake and Callaghan, 2008c– Chapter #6), and neutrally flexed specimens exposed to 1.5° of rotation for 10,000 cycles reported in the literature. Without flexion, the failure patterns were reported to occur in the endplates, facets, laminae and capsular ligaments, but not the disc. However, with flexion the repetitive axial twist rotational displacements caused damage primarily to the disc. If the load distribution was unchanged, the higher axial rotation angle should have caused the specimen to fail in less cycles of loading, and the failure pattern should not have changed. Modulators of this hypothesized mechanism include the velocity of the applied twist moment and the effects these have on the failure parameters and injury outcomes. The three physiologic loading rates investigated in this work were not shown to affect the ultimate axial twist rotational failure angle or moment in a neutral flexion/extension posture, but were shown to modify flexion-extension stiffness (Drake and Callaghan, 2008b– Chapter #5). All of the flexion-extension stiffness values post failure, from a one-time axial twist exposure, was less than those from a repetitive combined loading exposure that has been established to damage the intervertebral disc but not the facets. Therefore, it is likely that the facet joint provides the primary resistance to acute axial twist moments when the spine is in a neutral flexion posture, but there appears to be a redistribution of the applied load from the facets to the disc in repetitive exposures. The aforementioned studies determined there are changes in load distribution and load response caused by altered mechanics resulting from twist loading, but whether the exposures could possibly produce pain needed to be addressed. Previous research has determined that the disc has relatively low innervation in comparison to the richly innervated facet capsule and vertebra, with only the outer regions being innervated. Likewise, it is assumed that pain could be directly generated as the nucleus pulposus disrupted the innervated outer annular fibres in the process of herniation. Also, direct compression of the spinal cord or nerve roots has been shown to occur from the extruded nucleus and result in the generation of pain responses. Additionally, the nucleus pulposus has been shown to be a noxious stimulus that damages the function and structure of nerves on contact. The other source of nerve root compression commonly recognized is a decrease in intervertebral foramina space, which was previously believed to only be caused through losses in disc height. However, decreased intervertebral foramina space due to repetitive motions appears to be a viable pain generating pathway that may not directly correspond to simply a loss of specimen or disc height (Drake and Callaghan, 2008c– Chapter #6). This is new evidence for combined loading to generate pain through spinal deformation. The objective of many traditional treatments for nerve root compression focus on restoring lost disc height to remove the nerve root compression. Unfortunately, nerve root compression caused by repetitive loading may not be alleviated through this approach. This collection of studies was focused on determining whether altered load distribution in the spine, specifically between the intervertebral disc and facets, in response to applied axial twist loading (when added in combination with one and two axes of additional loading) was occurring, and examining how these modes of loading can contribute and/or alter the development of injury and pain. Therefore, findings generated from this thesis may have important implications for clinicians, researchers, and ergonomists.

Continental glaciers covered as much as thirty percent of the present-day inhabited earth during the Quaternary period. Traditionally, one-dimensional consolidation has been considered as the main process of formation for the soils deposited during glaciation. One of the outcomes of accepting one-dimensional consolidation as the main process of formation is that the geomechanical properties of soil in a horizontal plane are isotropic (known as cross-anisotropy). Recent measurements of subglacial pore pressure and preconsolidation pressure profile have indicated that this might not be the case. The role of subglacial shear action has probably been long neglected. The main objective of this research is to investigate the effects of subglacial shearing on the geomechanical properties of glaciated soils. <p> Recent research has found evidence of horizontal property anisotropy associated with the direction of the ice-sheet movement. A testing program was thus proposed to explore the relationship between the anisotropy of property and the direction of past glacier movement. The program involves several fundamental engineering parameters of soils. These parameters together with the corresponding test methods are as follows: (i) Conventional oedometer test yield stress anisotropy; (ii) Oedometer test with lateral stress measurement stiffness anisotropy; (iii) Load cell pressuremeter (LCPM) test in situ stress anisotropy. <p> The physical meaning of yield stress determined by conventional oedometer tests was interpreted as the critical state of structural collapse. The literature review and an experimental study on kaolin samples with a known stress history suggested that yield stress possesses certain dependency on the sampling direction. The anisotropy of yield stress for Battleford till from Birsay, Saskatchewan was also explored by testing directional oedometer samples. In addition, the anisotropy of stiffness was also investigated using a newly developed lateral stress oedometer that is capable of independent measurement of horizontal stresses at three different points with angles of 120 degrees. Preliminary evidence of a correlation between the direction of maximum stiffness in a horizontal plane and the known direction of glacial shear was observed. The correlation between the direction of maximum yield stress and known direction of glaciation was rather poor. Anisotropy of in situ stresses was investigated by conducting LCPM tests in Pot clay in the Netherlands. Based on the LCPM test results, it was concluded that the evidence of a correlation between the anisotropy of in situ stress and known direction of glacial advance is still rather obscure. <p> Although both the laboratory studies and field studies cannot sufficiently confirm the existence of lateral anisotropy of geomechanical properties and its relationship to the direction of the Quaternary ice-sheet movement, the effects of subglacial shearing should not be neglected in assessing the geotechnical properties of glaciated soils. In practice, it is usually found that the preconsolidation pressure profile does not follow the gravitational line as predicted by the one-dimensional consolidation theory and its magnitude is not compatible with the measured effective pressure values at the base of the glacier. It has been suggested that changes in seepage gradient (upward or downward) are responsible for the deviation of preconsolidation pressure profile away from the gravitational line. In this thesis, a new glacial process model consolidation coupled shearing was proposed. This model is based on the framework of traditional soil mechanics (critical state theory, Modified Cam-clay model and one-dimensional consolidation theory) and is consistent with the general geological and glaciological evidences. This model may provide an alternative explanation for the preconsolidation pressure patterns generally observed in practice. It can also be combined with groundwater flow characteristics to explain the diversity of the preconsolidation consolidation patterns. The proposed model was used successfully to obtain the preconsolidation pressure profile observed in Battleford till at Birsay and the subglacial shear-softening phenomenon.

Biological systems are able to perform complex movements with high energy-efficiency and, in general, can adapt to environmental changes more elegantly than traditionally engineered mechanical systems. The Equilibrium Point Hypothesis describes animal motor control as trajectories of equilibrium joint angle and joint stiffness. Traditional approaches to robot design are unable to implement this control scheme because they lack joint actuation methods that can control mechanical stiffness, and, in general, they are unable to take advantage of energy introduced into the system by the environment. In this paper, we describe the development and implementation of an FPGA-controlled, servo-actuated robotic joint that incorporates series-elastic actuation with specially developed nonlinear springs. We show that the joint's equilibrium angle and stiffness are independently controllable and that their independence is not lost in the presence of external joint torques. This approach to joint control emulates the behavior of antagonistic muscles, and thus produces a mechanical system that demonstrates biological similarity both in its observable output and in its method of control.

The main topic of this thesis is the performance improvement of microresonators as mass-sensitive biochemical sensors in a liquid environment. Resonant microstructures fabricated on silicon substrates with CMOS-compatible micromachining techniques are mainly investigated. Two particular approaches have been chosen to improve the resolution of resonant chemical/biochemical sensors. The first approach is based on designing a microresonator with high Q-factor in air and in liquid, thus, improving its frequency resolution. The second approach is based on minimizing the frequency drift of microresonators by compensating for temperature-induced frequency variations. A disk-shape resonant microstructure vibrating in a rotational in-plane mode has been designed, fabricated and extensively characterized both in air and in water. The designed resonators have typical resonance frequencies between 300 and 1,000kHz and feature on-chip electrothermal excitation elements and a piezoresistive Wheatstone-bridge for vibration detection. By shearing the surrounding fluid instead of compressing it, damping is reduced and quality factors up to 5800 in air and 94 in water have been achieved. Short-term frequency stabilities obtained from Allan-variance measurements with 1-sec gate time are as low as 1.1 10-8 in air and 2.3 10-6 in water. The performance of the designed resonator as a biological sensor in liquid environment has been demonstrated experimentally using the specific binding of anti-beta-galactosidase antibody to beta-galactosidase enzyme covalently immobilized on the resonator surface. An analytical model of the disk resonator, represented by a simple harmonic oscillator, has been derived and compared with experimental results. The resonance frequency and the Q-factor of the disk resonator are determined from analytical expressions for the rotational spring constant, rotational moment of inertia, and energy loss by viscous damping. The developed analytical models show a good agreement with FEM simulation and experimental results and facilitate the geometrical optimization of the disk-type resonators. Finally, a new strategy to compensate for temperature-induced frequency drifts of resonant microstructures has been developed based on a controlled stiffness modulation by an electronic feedback loop. The developed method is experimentally verified by compensating for temperature-induced frequency fluctuations of a microresonator. In principle, the proposed method is applicable to all resonant microstructures featuring excitation and detection elements.

Deterioration in the mechanical properties of concrete, masonry and steel structures are usually observed under repeated cyclic loading in the inelastic response range. Therefore such a behavior becomes critical when these types of structures are subjected to ground motions with specific characteristics. The objective of this study is to address the influence of degrading behavior on simple systems. The Structural Performance Database on the PEER web site, which contains the results of cyclic, lateral-load tests of reinforced concrete columns, are employed to quantify the degradation characteristics of simple systems by calibrating the selected degrading model parameters for unloading stiffness, strength and pinching of a previously developed hysteresis model. The obtained values of parameters from cyclic test results are compared with the recommended values in literature. In the last part of the study, response of SDOF systems with various degradation characteristics are investigated using a set of seismic excitations recorded during some major earthquakes. The results indicate that when all the degradation components are combined in a structural system, the effect of degradation on response values becomes much more pronounced.

In this study, the effect of pipe-soil relative stiffness on the behaviour of buried flexible pipes was investigated considering the pipe size, material type, stiffness, pipe-soil and natural soil-backfill interfaces and geometry of the trench using the finite element method. For this purpose, a parametric study was conducted to examine the effect of different variables on the resulting earth loads and deformations imposed on the buried pipes. Various types of trench pipe-soil cases were analysed for a certain natural ground and backfill material by the PLAXIS finite element code which allows simulating non-linear soil behaviour, the stages of construction as well as the pipe-soil interaction aspects of the problem. Loads and deformations obtained by the finite element method were compared with those calculated by the conventional approaches for different pipe-soil stiffness ratios. The finite element results obtained for the deformation of typically flexible Polyethylene pipes were then used to back-calculate the range of modulus of soil reaction, E&#039 / , values for various pipe-soil relative stiffness and they were compared with the suggested value proposed by Howard (1977).

The Savannah River Site (SRS) deposits in the Southeastern US between 30-45 m of depth are calcium carbonate-rich, marine-skeletal, Eocene-aged sediments with varying clastic content and extensive diagenetic alteration, including meter-sized caves that coexist with brittle and hard limestone. An experimental investigation including geotechnical (P- and S-wave velocities, tensile strength, porosity) and geochemical (EDS, XRD, SEM, N2-adsorption, stable isotopes, K-Ar age dating, ICP-assisted solubility, groundwater) studies highlighted the contrast between hard and brittle limestones, their relationship with cave formation, and allowed calculation of parameters for geochemical modeling. Results demonstrate that brittle and hard limestones bear distinct geochemical signatures whereby the latter exhibits higher crystallinity, lower clastic load, and freshwater-influenced composition. Results also reveal carbonate diagenesis pathways likely driven by geologic-time seawater/freshwater cycles, microorganism-driven micritization, and freshwater micrite lithification. The second section of this investigation dealt with SRS surface soils which are largely coarse-grained and rich in iron oxides with various degrees of maturity. These soils were simulated in the laboratory using Ottawa sands that were chemically coated with goethite and hematite. Surface (SEM, AFM, N2-adsorption) and geotechnical properties (fabric, small-strain stiffness, shear strength) were investigated on the resulting "soil analog". Results indicate that iron-oxide coated sands bear distinct inherent fabric and enhanced small-strain stiffness and critical state parameters when compared to uncoated sands. Contact mechanics analyses suggest that iron oxide coatings yield an increased number of grain-to-grain contacts, higher surface roughness, and interlocking, which are believed to be responsible for the observed properties.