Dissertations / Theses on the topic 'Lateral spread'

Our knowledge of the genetic contribution to Amyotrophic Lateral Sclerosis (ALS) is rapidly growing, and there is increasing research into how ALS spreads through the motor system and beyond. This thesis examines how genetic and non-genetic factors in ALS influence its spread. The genetic methods employed were PCR, genotyping, AFLP, DNA sequencing and gene expression. The methods to examine spread were H&E staining, clinical history, age of onset (AOO), survival and health utility. Statistical procedures applied included regression analyses of genetic and non-genetic factors, maximum likelihood estimation of genetic-phenotype variance and health utility, RNA-sequence analyses, and differential gene expression analysis. I found (A) that variation in the ATXN2 gene contributes to ALS, as does variation in C9ORF72 after correcting for the known C9ORF72 pathological hexanucleotide repeat (HREM). (B) On comparing regions of the spinal cord, patterns of differential gene expression between ALS cases and controls appeared consistent with spread and pathology. Functional annotation clustering revealed these genes were mostly involved in blood vessel and angiogenin-like function, glycoprotein-based activity, and leukocyte activity. (C) A significant proportion of survival variance in ALS could be explained by genetic variance. There were SNPs that predicted survival and AOO, one showing epistasis with the C9ORF72 HREM. (D) When modelling ALS progression using staging and a clinical trial dataset, the time and duration of each stage was statistically predictable. Staging also predicted health utility and other functional and psycho- metrics. The spread and pathology in ALS spinal cord regions affected gene expression profiles, which is likely a consequence of genetic susceptibility in that region. Indicators of spread, AOO and survival, could be predicted using genotypes. Disease progression was predictable as measured by clinical staging and health metrics. In summary, ALS spread seems to occur at a fixed rate in an individual and is influenced by genetics.

Characterization of the seismic hazard and ground-failure hazard of a site using traditional empirical lateral spread displacement models requires consideration of uncertainties in seismic loading, site conditions, and model prediction. Researchers have developed performance-based design methods to simultaneously account for these sources of uncertainty through the incorporation of a probabilistic analytical framework. While these methods can effectively handle the various sources of uncertainty associated with empirical lateral spread displacement prediction, they can be difficult for engineers to perform in a practical manner without the use of specialized numerical tools. To make the benefits of a performance-based approach accessible to a broader audience of geotechnical engineers, a simplified performance-based procedure is introduced in this paper. This map-based procedure utilizes a reference soil profile to provide hazard-targeted reference displacements across a geographic area. Equations are provided for engineers to correct those reference displacements for site-specific soil conditions and surface geometry to produce site-specific, hazard-targeted estimates of lateral spread displacement. The simplified performance-based procedure is validated through a comparative study assessing probabilistic lateral spread displacements across several cities in the United States. Results show that the simplified procedure closely approximates the results from the full performance-based model for all sites. Comparison with deterministic analyses are presented, and the place for both in engineering practice are discussed.

A full-scale ASTM E 1321 Lateral ignition and Flame Transport (LIFT) apparatus was constructed and compared with a Reduced scale Ignition and Flame spread Test apparatus (RIFT) adaptation of the cone calorimeter in the vertical position. The objective was to find a low cost and simple alternative to the LIFT apparatus for measuring opposed flow flame spread. Ignition tests were conducted using the LIFT, RIFT and ISO 5657 ignitability apparatus and flame spread experiments were conducted in the LIFT and RIFT. Nine different types of timber based products were tested for ignition and flame spread, and Quintiere's flame spread model was applied to the results to obtain material properties, such as thermal inertia, flame spread parameter and the minimum heat flux required for flame spread. These materials included plywood, medium density fibreboard (MDF), hardboard, particle board flooring, Melamine (Melteca) covered MDF, New Zealand Rimu, and Beech and New Zealand grown Macrocarpa and Radiata (Monterey) Pine. Further limited tests were conducted on Melteca covered particle board, and a second brand of particle board. The materials in the RIFT were tried with and without preheating to equilibrium. In addition, a view factor for the RIFT was developed, based on earlier work for the cone calorimeter element. The view factor equation was experimentally tested against measured values, and the calculated value was consistently lower than the experimental values, with similar flux profile. The standard procedure is for the material to be preheated before ignition, so that the surface is at equilibrium. The spread of the flame front past points on the sample surface after ignition is recorded, and from the flame front velocity and the model by Quintiere, material specific properties can be derived. The lack of preheating was found to affect the final results, by reducing the flame spread velocity and increasing the scatter in the experimental results. The RIFT gives comparable results to the same materials tested in the LIFT and to the published literature. The results the flame spread parameter and the minimum flux for flame spread are usually higher for the RIFT against the same material in the LIFT. There proved to be an effective limit on suitable materials able to be successfully tested in the RIFT to those that have a minimum flux for flame spread of less than 7kW/m2, with this limitation is dictated by the flux profile along the sample, and the lower resolution dictated by the smaller size. It is approximately equivalent to a minimum ignition flux of 18kW/m2.

This study set out to examine the effectiveness and reliability of six modifications to the Zhang et al. (2004) CPT-based lateral spread model. A regression analysis, distribution charts, and a discriminant analysis are performed to determine how effective the modifications are on the model. From the comparisons and statistical analysis performed in this study, application of these modifications reduces over-predictions from strain-based prediction methods. Unfortunately, the tendency to under-predict displacements on average is also increased.

Improving seismic hazard analysis is an important part of building safer structures and protecting lives. Since large magnitude earthquakes are rarer than other earthquakes, it is harder to model seismic hazards such as lateral spread displacements for these events. Engineers are often required to extrapolate current lateral spreading models when designing utilities, bridges, and piers to withstand the ground displacements caused by earthquakes with magnitudes larger than 8.0. This study uses three case histories from the Maule Chile 2010 earthquake (Mw =8.8) to develop recommendations on which models are most accurate for large earthquake events and how to improve the accuracy of the models. Six empirical models commonly used in engineering practice are compared. The model that best matches the Maule Chile case histories uses local attenuation relationships to make it easier to apply the model to any seismic region. Models that use lab data from cyclic shear tests over predict displacements but using a strain-reduction factor with depth significantly improved the accuracy of the results. Site-to-source distances can vary greatly between geographic seismic and faulting mechanisms. For this reason, models that depend on an internal source-to-site distance show less promise with large subduction zone earthquakes throughout the world. Models with site-to-source distances are most accurate in the western United States and Japan because the case histories for these models came from those countries.

Liquefaction-induced lateral spread displacements cause severe damage to infrastructure, resulting in large economic losses in affected regions. Predicting lateral spread displacements is an important aspect in any seismic analysis and design, and many different methods have been developed to accurately estimate these displacements. However, the inherent uncertainty in predicting seismic events, including the extent of liquefaction and its effects, makes it difficult to accurately estimate lateral spread displacements. Current conventional methods of predicting lateral spread displacements do not completely account for uncertainty, unlike a performance-based earthquake engineering (PBEE) approach that accounts for the all inherent uncertainty in seismic design. The PBEE approach incorporates complex probability theory throughout all aspects of estimating liquefaction-induced lateral spread displacements. A new fully-probabilistic PBEE method, based on results from the cone penetration test (CPT), was created for estimating lateral spread displacements using two different liquefaction triggering procedures. To accommodate the complexity of all probabilistic calculations, a new seismic hazard analysis tool, CPTLiquefY, was developed. Calculated lateral spread displacements using the new fully-probabilistic method were compared to estimated displacements using conventional methods. These comparisons were performed across 20 different CPT profiles and 10 cities of varying seismicity. The results of this comparison show that the conventional procedures of estimating lateral spread displacements are sufficient for areas of low seismicity and for lower return periods. However, by not accounting for all uncertainties, the conventional methods under-predict lateral spread displacements in areas of higher seismicity. This is cause for concern as it indicates that engineers in industry using the conventional methods are likely under-designing structures to resist lateral spread displacements for larger seismic events.

Over the past several decades, empirical formulas have been developed and improved to predict liquefaction and lateral spread based on a database of case histories from observed earthquakes, such as Youd et al. (2002) and Rauch and Martin (2000). The 2010 Maule Chile earthquake is unique first of all because it is recent and was not used to develop recent liquefaction and lateral spread evaluation methods, and therefore can be reasonably used to evaluate the effectiveness of such equations. Additionally, the 8.8 magnitude megathrust event fills a significant gap in the databases used to develop these empirical formulas, which tends to under represent large magnitude earthquakes and events which occur along subduction zones. Use of case histories from this event will therefore effectively test the robustness and accuracy of these methods.As a part of this comparison, data will be collected from two piers in Port Coronel, Chile: Lo Rojas or Fisherman's Pier, and el Carbonero. Lo Rojas is a municipally owned pier which failed in the 2010 earthquake. Dr. Kyle Rollins gathered detailed engineering survey data defining lateral spread displacements along this pier in a reconnaissance visit with other GEER investigators after the earthquake. El Carbonero was under construction during the earthquake, but no known lateral displacements were observed. Collaboration with local universities and personnel contributed a great deal of knowledge about the soil profile. In early April 2014, collection of SPT and CPT data began in strategic locations to fill gaps of understanding about the stratigraphy near the two piers. Additional testing will provide necessary information to carry out predictions of displacements using current empirical models, which can then be compared with observed displacements collected after the earthquake. Collected data will also be complied, and this alone will provide useful information as it represents a unique case history for future evaluation.The goals of this study are therefore: (1) Collect data for two piers (Lo Rojas and el Carbonero) in Port Coronel, Chile to provide a useful case history of lateral displacements observed; (2) Conduct a liquefaction and lateral spread analysis to predict displacement of the two piers in question, considering lateral spread and slope stability; (3) Compare predicted values with observed displacements and draw conclusions on the predictive capabilities of analyzed empirical equations for similar earthquakes (4) Make recommendations to improve when possible.

Shallow embedded column connections are widely used for columns resisting gravity loads in current design methods. These connections are usually considered “pinned” for structural analysis. In reality these connections fall in between a fixed and a pinned condition. Although methods exist to estimate the stiffness and strength of exposed columns or embedded columns under lateral loads, little research has been done to determine the strength of shallow embedded columns. An experimental study was carried out to investigate the strength of these connections. A total of 12 specimens with varying orientation, embedment depth, and column size were loaded laterally until failure or significant loss in strength. The results showed that shallow embedded connections are 86%-144% stronger in yielding and 32%-64% stronger in ultimate strength than current design methods would predict. This strength comes from a combination of the embedment depth and the resistance from the base plate and anchor rods. A model is proposed to explain the strength of the specimens and to conservatively estimate the strength of specimens with different variables. The specimens also exhibited stiffness ranging from 50%-75% of what would be expected from fully embedded columns.

The 2010, moment magnitude 8.8 earthquake that occurred near Maule, Chile caused major damages to several piers in the Port of Coronel located approximately 160 kilometers (100 miles) to the South of the earthquake epicenter. One of the piers, the North Pier, experienced significant lateral spreading that was caused from liquefaction of the soils at the approach zone of the pier. Damages from lateral spreading and liquefaction effects consisted of sheet pile welding ruptures of the cross-support beams, stiffener buckling, pile displacements, pile rotations, and pier deck displacement. Researchers analyzed the North Pier after the earthquake and documented in detail the damage caused by lateral spread displacements. This study introduces a simplified performance-based procedure called the "Simplified Modeling Procedure" that is used for the analysis of piles supporting a pier that are exposed to lateral spread displacements. The procedure uses the software LPILE, a common program for analyzing a single pile under lateral loading conditions, to evaluate a more complex multi-pile pier design. Instead of analyzing each of the piles in a given pier individually, the procedure utilizes what is known as a "Super Pile" approach to combine several piles into a single representative pile during the analysis. To ensure displacement compatibility between each "Super Pile" in the analysis, the "Super Piles" are assumed to be fully connected at the top of each "Super Pile" to the pier deck. The Simplified Modeling Procedure is developed and tested using the case study history of the North Pier from the Port of Coronel during the 2010 Maule earthquake. The Simplified Modeling Procedure incorporates p-y springs with a lateral push-over analysis. This approach allows the analysis to be performed in a matter of seconds and allows the user to more easily draw the needed correlations between the rows of piles. This procedure helps identify that different rows of piles either contribute to the movement of the pier or contribute to the bracing of the pier. The procedure ultimately predicts the anticipated pier deck deflection by determining when all the pile rows and their respective shear forces are in equilibrium. The Simplified Modeling Procedure predicted that the North Pier experienced deflections between approximately 0.31 meters (1.01 feet) and 0.38 meters (1.26 feet). The predicted deflections and rotations determined using the procedure were determined to be a relatively close representation of the observations made during the post-earthquake reconnaissance observations.

β-N-methylamino-L-alanine (BMAA) is one of the more recently discovered bioactive compounds produced by cyanobacteria. BMAA is a non-protein amino acid reported present in human brain tissues of patients deceased from a neurodegenerative disease, such as Alzheimer´s disease or amyotrophic lateral sclerosis (ALS). This observation in combination with its neurotoxic effects in eukaryotes (in vivo and in vitro) and its potential to incorporate into (human) proteins, causing protein aggregation, suggests BMAA as a possible causative environmental agent for neurodegenerative diseases. Due to the ubiquitous nature of cyanobacteria with a wide occurrence in both aquatic and terrestrial environments, BMAA could be globally spread. Hence, investigations of a possible coupling between BMAA and neurodegeneration are urgently needed as well as to identify sources of BMAA in Nature. The aim of this thesis was to examine the potential occurrence of BMAA in bloom forming cyanobacteria of the Baltic Sea and its possible transfer to other organisms of this ecosystem. Of importance was also to reveal any likely routes for human BMAA exposure in the Baltic Sea region and to further investigate BMAA as a triggering agent for neurodegenerative diseases. Acknowledged difficulties of analysing BMAA in biological samples also inferred method development as part of the experimental studies. Investigating the role of BMAA in its producers was another purpose of the thesis, which may be crucial for future management of BMAA-producing cyanobacteria. By screening natural populations of the major filamentous bloom forming cyanobacteria of the Baltic Sea, we discovered the presence of BMAA throughout the entire summer season of two consecutive years, using a highly specific analytical method (liquid chromatography-tandem mass spectrometry; LC-MS/MS). BMAA was found to bioaccumulate in zooplankton and fish, as well as in mussels and oysters from the Swedish west coast. To improve the understanding of BMAA analyses in natural samples, the formation of carbamate adducts in the presence of bicarbonate was examined. Using two derivatization techniques in combination with LC-MS/MS, we could show that BMAA detection was not hindered by carbamate formation. Exogenously added BMAA inhibited nitrogen fixation in the model cyanobacterium Nostoc sp. PCC 7120, which was also hampered in growth and displayed signs of nitrogen starvation. Finally, BMAA was detected in cerebrospinal fluid in three of 25 Swedish test individuals, and represents the first confirmation of BMAA in the human central nervous system using LC-MS/MS as the primary analytical method. However, the association of BMAA to neurodegenerative diseases could not be verified as BMAA was present in both control individuals (two) and in one ALS-patient. Nevertheless, the finding of a known neurotoxic compound in the human central nervous system is alarming and potential consequences should be investigated. The discovery of the neurotoxic compound BMAA in Baltic Sea organisms, and in the central nervous system of humans potentially consuming fish from this ecosystem is concerning and warrants continued investigations of BMAA occurrence and human exposure. Further knowledge on the function and regulation of BMAA may help in developing strategies aiming to minimise human exposure.

At the time of the doctoral defense, the following paper was unpublished and had a status as follows: Paper 3: Manuscript.

Lateral spread is a seismic hazard associated with soil liquefaction in which permanent deformations are developed within the soil profile due to cyclic mobility. Lateral spread has historically been one of the largest causes of earthquake-related damage to infrastructure. One of the infrastructure components most at risk from lateral spread is that of deep foundations. Because performance-based engineering is increasingly becoming adopted in earthquake engineering practice, it would be beneficial for engineers and researchers to have a performance-based methodology for computing pile performance during a lateral spread event. This study utilizes the probabilistic performance-based framework developed by the Pacific Earthquake Engineering Research Center to develop a methodology for computing probabilistic estimates of kinematic pile response. The methodology combines procedures familiar to most practicing engineers such as probabilistic seismic hazard analysis, empirical compution of lateral spread displacement, and kinematic pile response using p-y soil spring models (i.e. LPILE). The performance-based kinematic pile response model is applied to a series of lateral spread case histories from the earthquake that struck the Limon province of Costa Rica on April 22, 1991. The M7.6 earthquake killed 53 people, injured another 193 people, and disrupted an estimated 30-percent of the highway pavement and railways in the region due to fissures, scarps, and soil settlements resulting from liquefaction. Significant lateral spread was observed at bridge sites throughout the eastern part of Costa Rica near Limon, and the observed structural damage ranged from moderate to severe. This study identified five such bridges where damage due to lateral spread was observed following the earthquake. A geotechnical investigation is performed at each of these five bridges in an attempt to back-analyze the soil conditions leading to the liquefaction and lateral spread observed during the 1991 earthquake, and each of the five resulting case histories is developed and summarized. The results of this study should make a valuable contribution to the field of earthquake hazard reduction because they will introduce a procedure which will allow engineers and owners to objectively evaluate the performance of their deep foundation systems exposed to kinematic lateral spread loads corresponding to a given level of risk.

Shallowly embedded column base connections with unreinforced block out concrete are a common method of connecting steel columns to their foundation. There has been little research done to accurately quantify the effects of this block out concrete on the connection strength and rigidity, and therefore there is nothing to aid the practicing engineer in accounting for this in structural analysis. Due to this lack of understanding, engineers have typically ignored the effects of shallow block out concrete in their analysis, presumably leading to a conservative design. Recent research has attempted to fill this gap in understanding. Several methods have been proposed that seek to quantify the effects of shallow block out concrete on a column base connection. Barnwell proposed a model that predicts the strength of a connection. Both Jones and Tryon used numerical modeling to predict the rotational stiffness of the connection. An experimental study was carried out to investigate the validity of these proposed models. A total of 8 test specimens were created at 2/3 scale with varying column sizes, connection details, and embedment depths. The columns were loaded laterally and cyclically at increasing displacements until the connection failed. The results show that the strength model proposed by Barnwell is reasonable and appropriate, and when applied to this series of physical tests produce predictions that have an observed/predicted ratio of between 0.95 to 1.39. The results also show that methods for estimating the rotational stiffness of the connection at the top of the block out concrete, as proposed by Jones and Tryon also produce reasonable values that had observed/predicted ratios of between 0.93 to 1.47. An alternative model for determining a design value for the rotational stiffness of a shallowly embedded column base plate is also proposed. When the embedment depth to column depth ratio is greater than 1.22, the connection is sufficiently rigid and at small deflections (less than 1% story drift) may be accurately modelled with infinite rotational stiffness (a "fixed" connection) at the base of the column.

La mise en œuvre de méthodes d'assimilation d'ensemble est une technique assez récente visant à simuler les erreurs d'analyse et de prévision d'un système d'assimilation de données. Cela permet d'une part d'estimer des covariances spatiales des erreurs de prévision, qui sont un ingrédient essentiel des systèmes d'assimilation de données, dans la mesure où elles permettent de filtrer et de propager spatialement l'information observée. La dépendance de ces covariances d'erreur à la situation météorologique devient ainsi accessible avec ces techniques d'ensemble. D'autre part, l'assimilation d'ensemble est également une méthode de plus en plus utilisée pour fournir des perturbations initiales aux systèmes de prévision d'ensemble. Une telle approche peut être mise en place non seulement dans un système modélisant l'atmosphère sur l'ensemble du globe, mais aussi dans un système régional à aire limitée, en utilisant dans ce cas des conditions limites latérales appropriées. Le sujet de thèse proposé consiste à examiner certaines propriétés de sensibilité de ces techniques d'assimilation d'ensemble dans ces deux types de contextes (à savoir global et régional, respectivement). Il s'agit premièrement d'étudier la sensibilité d'un système global d'assimilation d'ensemble à son initialisation. Cela sera mené en comparant une technique d'initialisation "à froid" (basée sur des perturbations initiales nulles) avec une méthode basée sur des perturbations initiales tirées d'un modèle de covariance. Dans une deuxième partie, la sensibilité d'une assimilation d'ensemble régionale aux conditions limites latérales sera examinée. Dans cette perspective, une comparaison entre différentes techniques de production des perturbations latérales sera réalisée. Il s'agit notamment de comparer les approches basées sur des perturbations latérales qui sont soit nulles, soit tirées d'un ensemble global, ou encore produites à l'aide d'un modèle de covariance. Ces études de sensibilité seront menées d'une part en utilisant des expérimentations avec les systèmes global Arpege et régional Aladin. Ce travail s'appuiera d'autre part sur une formalisation des équations qui gouvernent l'évolution des perturbations au sein d'une assimilation d'ensemble. Ces études devraient permettre de documenter les propriétés de ces assimilations d'ensemble, et de définir des stratégies de mise en œuvre en grandeur réelle pour l'assimilation de données ainsi qu'éventuellement pour la prévision d'ensemble.

碩士
國立中正大學
機械工程系研究所
107
During the hot rolling process of aluminum slab, the slab is rolled by the work roll to reduce the thickness. The deformation in the width direction is called the slab lateral spread. Since the surface of the slab is in contact with the work roll, the cooling rate at the slab surface is different from that in the slab center. The temperature gradient causes difference in the deformation resistance of the slab along the thickness direction, resulting in uneven deformation of the side wall to form a concave shape, which is called lateral spread. After multi-pass rolling, edge lamination occurs on both sides of the slab. The edge lamination on both sides must be removed by trimming machine, which reduces the yield rate and raises the cost. The temperature gradient in the thickness of the slab directly affects the level of the lateral spread. In this study, a three-dimensional rigid-plastic rolling model was established based on the software ABAQUS, and the multi-pass thermal-mechanical coupled simulation was carried out. According to the operating conditions set by the experimental mill, the lateral spread geometric shape, rolling force, rolling torque, temperature distribution in the thickness of the slab were predicted by the finite element model. The finite element simulation results were compared with those obatained by experiments. The simulation model was then tuned according to the experimental results until prediction of the lateral spread is verified, the goal of the work is to provide a numerical prediction tool to effectively reduce the lateral spread of the slab to reduce or to eliminate the edge lamination during that rolling of aluminum slab.

博士
國立成功大學
生物醫學工程學系
105
Purpose: Identification of positive lateral retropharyngeal lymph (LRPL) nodes in nasopharyngeal carcinoma (NPC) is important. The primary objective of this study is to improve the diagnostic accuracy of LRPL node (LRPLN) with three parts. Part I is by determining the size range where the recovery coefficient (RC) method of fluorodeoxyglucose (FDG) positron emission tomography-computed tomography (PET/CT) be helpful in detecting LRPL nodal metastases of NPC patients previously treated with radiation therapy. In part II, we use neural network (NN) to diagnose of LRPLNs automatically with parameters in PET/CT and magnetic resonance imaging (MRI). In part III, we explore other nodal parameters in MRI and PET/CT for increasing the prediction accuracy although minimal axial diameter (MIAD) in magnetic resonance imaging (MRI) has been recognized as the most useful parameter in diagnosing LRPLNs in NPC. Multi-stage approach with new diagnostic criterion for better accuracy and clinically was assessed. Materials and Methods: The group I patient, a total of 142 LRPLNs assessed by MRI in 71 NPC patients was retrospectively chosen for partial volume correction study. LRPLNs with central necrosis, extracapsular invasion, or asymmetric grouping or those ascertained on follow-up MRI scans were considered positive for metastases. The criterion for positive diagnosis of nodal metastasis on FDG PET/CT scans was defined as maximal standard uptake value (SUVmax)  2.5. Nodes not separated from main tumors were excluded. An established RC method, the sphere-to-background ratio, was employed. Nodes were further categorized into three groups of minimal axial diameters: below 6 mm, 6 to 6.9 mm, and above 7 mm. There was group II patient with totally 148 patients and 269 LRPLNs from China. These LRPLNs ascertained on follow-up MRI were considered positive for metastases. In part II and part III, both groups of patients and LPRLNs were used and a total of 411 LRPLNs were retrospectively collected from 219 patients with NPC. NN model was tested for 15 combinations of four parameters, namely MIAD, maximal axial diameter (MAAD), and maximal coronal diameter (MACD) and mean standard uptake value (NSUVmean). The optimal cutoff value of each parameter was derived for each parameter. The results of NN were compared with expert evaluation from 142 LRPLNs. The multi-stage approach for new criterion determined through was accessed through modified exhaustive search, and the new criteria were compared to single MIAD using a bootstrap sampling method. Results: With RC method, a total of 88 separable LRPLNs were examined by FDG PET. Thirty-five nodes were positive and fifty-three nodes were negative. The RC method significantly improved sensitivity (from 20% to 100%) and accuracy (from 14% to 71%) for nodes sized 6 to 7 mm. The accuracy was improved from 92% to 96% for nodal size above 7 mm. In NN and multi-step approach for new criterion, the accuracy rate (percentage) for the MIAD is 366/411 (89.1%). The optimal cutoff value is 6.0 mm. With four parameters, the accuracy rate was 89.09% for NN evaluation and for expert evaluation, 115/142 (81.0%), respectively. In predication, the optimal combinations of four parameters resulted in correct identification six additional nodes (372/411, 90.5%), representing 13.3% (6/45) decreases in incorrect prediction, respectively. Conclusion: NPC LRPLNs with an MIAD ≥ 6.1 mm are positive. Among nodes with an MIAD 〈 6.1 mm, if the NSUVmean ≥ 2.6 or MACD ≥ 25 mm and MAAD ≥ 8 mm, the nodes are positive; otherwise, they are negative. Partial volume correction in PET/CT enhances the accuracy of detecting nodes in the size range of above 6 mm for LRPL nodal metastases of NPC. NN provides accurate and consistent diagnosis of LRPLNs comparable to expert judgment.

碩士
淡江大學
土木工程學系碩士班
97
This study following the methodology suggested by JRA, is to study the single pile dynamic behavior under lateral spreading of soil. It applies the programming language of Fortran to exploit the analysis. Liquefaction potential analysis in this study is Tokimatsu and Yoshimi method, which applies by Microsoft Excel. The research result is essentially founded on the model of wave equation analysis, and the difference equations are used to solve the problem. The stiffness of piles are started iterative analysis by using simplified Bouc-Wen model. This study aims to examine the deformation and destruction of the pile foundation. Three research results are derived in this conclusion: First, this study suggests adopting Ishihara’s theory of lateral earth pressure, in replace of the traditional fixed head, into consideration in order to the flexibility of non-liquefiable and base layer. Second, the data of the liquefaction potential index and the depth of liquefiable layer by calculating can be gathered by Liquefaction potential analysis. The liquefaction potential index is one of the JRA method parameters, and it can influence the earth pressure of the layer which is upon the liquefiable layer very much. The result of this study proves that the depth of liquefiable layer influences the behavior of pile foundation a lot. Third, the head of pile in this study is fixed head. The max displacement of pile occurs on the head of pile foundation. The larger value of the bending moment of pile occurs on the head of pile foundation or near the base of the liquefiable layer. According to the result of this study, the layer liquefies and make the soil flow when the earthquake is happened. The result of this study make us know well that the displacement and the deformation of pile foundation, and can consider that the destruction owing to the bending moment of pile foundation which is near the base of the liquefiable layer.

Drainage methods for liquefaction remediation have been in use since the 1970's and have traditionally included stone columns, gravel drains, and more recently prefabricated vertical drains. The traditional drainage techniques such as stone columns and gravel drains rely upon a combination of drainage and densification to mitigate liquefaction and thus, the improvement observed as a result of these techniques cannot be ascribed solely to drainage. Therefore, uncertainty exists as to the effectiveness of pure drainage, and there is some hesitancy among engineers to use newer drainage methods such as prefabricated vertical drains, which rely primarily on drainage rather than the combination of drainage and densification. Additionally, the design methods for prefabricated vertical drains are based on the design methods developed for stone columns and gravel drains even though the primary mechanisms for remediation are not the same. The objectives of this research are to use physical and numerical models to assess the effectiveness of drainage as a liquefaction remediation technique and to identify the controlling behavioral mechanisms that most influence the performance of sites treated with prefabricated vertical drains. In the first part of this research, a suite of three large-scale dynamic centrifuge tests of untreated and drain-treated sloping soil profiles was performed. Acceleration, pore pressure, and deformation data was used to evaluate the effectiveness of drainage in reducing liquefaction-induced lateral deformations. The results showed that the drains reduced the generated peak excess pore pressures and expedited the dissipated of pore water pressures both during and after shaking. The influence of the drains on the excess pore pressure response was found to be sensitive to the characteristics of the input motion. The drainage resulted in a 30 to 60% reduction in the horizontal deformations and a 20 to 60% reduction in the vertical settlements. In the second part of this research, the data and insights gained from the centrifuge tests was used to develop numerical models that can be used to investigate the factors that most influence the performance of untreated and drain-treated lateral spread sites. Finite element modeling was performed using the OpenSees platform. Three types of numerical models were developed - 2D infinite slope unit cell models of the area of influence around a single drain, 3D infinite slope unit cell models of the area of influence around a single drain, and a full 2D plane strain model of the centrifuge tests that included both the untreated and drain-treated slopes as well as the centrifuge container. There was a fairly good match between the experimental and simulated excess pore pressures. The unit cell models predicted larger horizontal deformations than were observed in the centrifuge tests because of the infinite slope geometry. Issues were identified with the constitutive model used to represent the liquefiable sand. These issues included a coefficient of volumetric compressibility that was too low and a sensitivity to low level accelerations when the stress path is near the failure surface. In the final part of this research, the simulated and experimental data was used to examine the relationship between the generated excess pore water pressures and the resulting horizontal deformations. It was found that the deformations are directly influenced by both the excess pore pressures and the intensity of shaking. There is an excess pore pressure threshold above which deformations begin to become significant. The horizontal deformations correlate well to the integral of the average excess pore pressure ratio-time history above this threshold. They also correlate well to the Arias intensity and cumulative absolute velocity intensity measures.
text

碩士
國立高雄第一科技大學
營建工程研究所
99
Pile foundation is the major public works and transportation construction commonly used in the world, in such as high-speed railway and highway, to withstand the transfer of lateral forces and axial forces. However, behaviors of piles in lateral spreading due to soil liquefaction is particular important and complicated. This paper researched responses of pile foundation in lateral spread based on a three dimensional soil–water coupled dynamic analysis. At first, numerical model of accordance with simulation to Abdoun et al. (2003) was to use physical single pile test in inclined ground of 50g centrifuge test. This part of study focused on responses of the single pile in two and three layered inclined ground with two liquefaction strength. And then, this study set up a model to simulate lateral responses on group pile foundation, referred to Uzuoka et al. (2008). Responses of the group piled foundation with some change of the height of footing, length of pile and in the layered of ground with different liquefaction strengths was discussed and compared with the one of single pile.

Liquefaction-induced deformations associated with lateral spreading represent a significant hazard that can cause substantial damage during earthquakes. The ability to accurately predict lateral-spreading displacement is hampered by a lack of field data from previous earthquakes. Remote sensing via optical image correlation can fill this gap and provide data regarding liquefaction-induced lateral spreading displacements. In this thesis, deformations from three earthquakes (2010 Darfield, February 2011 Christchurch, and 2011 Tohoku Earthquakes) are measured using optical image correlation applied to 0.5-m resolution satellite imagery. The resulting deformations from optical image correlation are compared to the geologic conditions, as well as field observations and measurements of liquefaction. Measurements from optical image correlation are found to have a precision within 0.40 m in all three cases, and results agree well with field measurements.
text

La sclérose latérale amyotrophique (SLA) est une pathologie complexe multifactorielle dont les mécanismes de dégénérescence des motoneurones et de propagation rapide au sein du système nerveux sont encore incertains. Par l’utilisation du nématode Caenorhabditis elegans, nous avons pu investiguer génétiquement et pharmacologiquement certains facteurs entrant en jeu dans la toxicité de TDP-43 et FUS. Des mutations dominantes dans ces protéines liant l’ADN et l’ARN, structurellement et fonctionnellement proches, sont des causes de SLA familiales. Nous avons, par le passé, construit un modèle de ver transgénique possédant le gène TARDBP ou FUS codant respectivement pour les protéines humaines TDP-43 ou FUS, sous le contrôle d’un promoteur exprimé seulement dans les neurones GABAergiques. Uniquement lorsque les gènes TARDBP ou FUS sont mutés, des symptômes relatifs à la SLA apparaissent au cours du temps, à savoir une paralysie progressive et une neurodégénérescence des motoneurones GABAergiques. Nous avons voulu connaître le rôle que pouvait jouer le système immunitaire, dont des évidences croissantes montrent une implication dans la SLA, dans la protéotoxicité liée à ces protéines dans nos modèles de ver. Dans un premier temps, nous avons évalué la motricité des vers en milieu solide et en milieu liquide, et grâce à des vers transgéniques exprimant la GFP dans les neurones GABAergiques, nous avons pu quantifier la neurodégénérescence. Nos résultats soulignent un rôle prévalent de l’orthologue de la protéine du système immunitaire innée Sarm1 chez le ver, TIR-1, ainsi que les kinases en aval dans la pathologie. Nous avons pu, de surcroît, utiliser le marqueur NLP-29 dont le promoteur lié à la GFP nous indique l’activation de la voie Sarm1 dans l’ensemble du ver et non seulement dans les neurones. De manière intéressante, l’activation de ces protéines se produit entre autres dans des cellules non-neuronales de manière paracrine suggérant qu’un signal de danger opère extracellulairement et vraisemblablement à travers un récepteur membranaire. Ces dernières années, un nombre important d’études met en lumière le rôle proéminent des microARNs dans des maladies telle que la SLA. Classiquement vus comme des régulateurs de l’expression post-transcriptionnelle, ce qui en font notamment des outils antiviraux puissants, ils peuvent agir à d’autres niveaux et notamment comme ligands de récepteurs Toll-like (TLRs), eux aussi impliqués dans la SLA. iv Outre le potentiel biomarqueur de ces petites molécules, nous avons investigué leur rôle dans la neurodégénérescence observée dans la SLA. Ainsi, dans une deuxième partie d’étude, nous avons utilisé des mutants pour différentes protéines impliquées dans la biogénèse des microARNs et trouvé qu’elles étaient partie intégrante du processus de paralysie et de dégénérescence des vers TDP-43A315T. Plus encore, le microARN let-7 pourrait être une molécule signal transitant entre les neurones et les cellules avoisinantes. Enfin, des analyses bio-statistiques prédisent la possibilité que let-7 se lie au récepteur TOL-1, l’unique orthologue des TLRs chez C. elegans. Les propriétés des microARNs en font en effet des cibles de choix dans la recherche de nouveaux acteurs dans la SLA et de potentielles cibles thérapeutiques.
Amyotrophic lateral sclerosis is a complex multifactorial pathology characterized by the progressive spread of motor neuron degeneration. Unfortunalety, the underlying disease mechanisms remain unclear. By using the nematode Caenorhabditis elegans, we were able to investigate genetically and pharmacologically some factors involved in TDP-43 or FUS proteotoxicity. Dominant mutations in these structurally and functionally similar DNA/RNA binding proteins, are causative for familial ALS. We have constructed transgenic C. elegans models expressing human TARDBP or FUS genes - encoding respectively TDP-43 and FUS - only in GABAergic motor neurons. In these transgenics animals, the expression of mutant TARDBP or FUS alleles results in early the motor deficits leading to age-dependent paralysis accompanied by neuronal protein aggregation. Using transgenic strain expressing GFP in GABAeric neurons, we found an increased rate of neurodegeneration in TDP-43 and FUS mutants. With these models we investigated the potential role of the innate immune system as a modifier of these phenotypes. Our results highlight a prevalent role for the worm’s innate immune system, and specifically the TIR-1/Sarm1 pathway and associated downstream kinases in neurodegeneration. We used GFP fluorescence linked to NLP-29 promoter to indicate Sarm1 pathway activation in the entire worm. Interestingly, activation of the TIR-1/Sarm1 pathway occurs in a paracrine manner in non-neuronal cells, suggesting that a danger signal operates extracellularly likely through a membrane receptor. In a past few years, a number of studies have highlighted the prominent role of microRNAs in diseases such as ALS. Traditionally seen as post-transcriptional regulators, what makes them powerful antiviral tools is that they can act at other levels and in particular as Toll-like receptors (TLRs) ligands, also involved in ALS. In addition to the biomarker potential of these small molecules, we investigated their role in the neurodegeneration observed in ALS. As a result, in the a second section of this study, we used worms mutant for several proteins involved in the biogenesis of microRNAs and found that they were involved in the process of TDP-43A315T-independent paralysis and neurodegeneration. Moreover, the microRNA let-7 seems to be a signal molecule involved in the non-cell autonomous trans-neuronal and trans-cellular spread of motor neuron degeneration. Finally, bio-statistical analyzes predict the possibility that let-7 binds to the vi TOL-1 receptor, the single ortholog of TLRs in C. elegans. Thus microRNAs may be prime targets for ALS therapeutic intervention.

In der vorliegenden Dissertation wurden die Folgen einer SPRED2-Defizienz in einem Knockout Mausmodell untersucht. Dabei wurde insbesondere die mögliche Verbindung zur Zwangsstörung, einer psychiatrischen Erkrankung beleuchtet. Das SPRED2-Protein kommt im menschlichen Körper in zahlreichen Geweben vor, besonders im Hirn wurde eine ubiquitäre Expression nachgewiesen und ein Zusammenhang mit der Neurogenese und neuronaler Differenzierung vermutet. Seine regulatorische Funktion besteht in einer inhibitorischen Wirkung auf den BDNF/TrkB-ERK-Signalweg, welcher u.a. für die Transkription neuronaler Gene verantwortlich ist. Die verwendeten SPRED2-defizienten Mäuse wurden durch Insertion eines Gene-Trap Vektors in das Spred2-Gen generiert. Die Insertion verhindert letztendlich die korrekte Translation des Proteins. Von der durch weitere Verpaarung entstehenden SPRED2-Knockout Mauslinie wurden ausschließlich männliche Tiere verwendet. Im Rahmen einer SPRED2-KO-Studie von der AG Schuh des Physiologischen Instituts der Universität Würzburg, die u.a. die Entgleisung der HHNA mit resultierendem erhöhten Stresshormonspiegel und eine Dysregulation des Mineralhaushaltshormons Aldosteron zeigte, wurden bei den Versuchstieren zwanghafte Verhaltensmuster beobachtet. Daraufhin wurden elektrophysiologische Messungen durchgeführt, die auf eine Anomalie in der synaptischen Übertragung zwischen Thalamus und Amygdala hindeuteten. Erhöhte Effizienz und Erregbarkeit der amygdaloiden Neuronen führten zu der morphologischen Untersuchung, die im Rahmen dieser Arbeit durchgeführt wurden. Da die Afferenzen des Thalamus vorwiegend in den lateralen Kern der Amygdala projizieren, wurde zunächst dieser betrachtet. Ziel der Untersuchung war es, Erkenntnisse darüber zu erlangen, ob der Knockout des SPRED2-Proteins in Mäusen zu einer veränderten Morphologie der Neuronen der lateralen Amygdala führt. Falls dies der Fall sein sollte, könnte damit zumindest ansatzweise das zwanghafte Verhalten der SPRED2-defizienten Mäusen erklärt werden. Die Hirne der Versuchstiere wurden nach der Golgi-Cox-Imprägnierung nach Glaser und Van der Loos und der Einbettung in Celloidin in 150 μm dicke Scheiben geschnitten und anschließend mithilfe eines Hellfeld-Mikroskops und des Neurolucida-Systems analysiert. Quantitativ erfasst und analysiert wurden pyramidale Klasse 1-Neuronen der lateralen Amygdala inklusive absoluter Anzahl und Dichte der Spines an ihren Dendriten. Die Untersuchung zeigte bei SPRED2-KO-Mäusen eine signifikante Erhöhung der mittleren Länge des apikalen Dendriten in Branch order 3 und eine tendenzielle Erhöhung der Gesamtzahl der Spines an den Dendriten in Branch order 1-3 gegenüber den Wildtyp-Mäusen. Daraus lässt sich folgern, dass ein Knockout des SPRED2-Proteins sich auf die Morphologie der Neuronen der lateralen Amygdala auswirkt. Die erhöhte mittlere Länge des apikalen Dendriten in Branch order 3 und die tendenziell erhöhte Spine-Anzahl korrelieren mit der gesteigerten synaptischen Übertragung und Erregbarkeit an amygdaloiden pyramidalen Neuronen. Auf molekularer Ebene kann die Hyperaktivität der lateralen Amygdala als Folge der fehlenden Inhibition des BDNF/TrkB-ERK-Signalwegs und der dadurch veränderten Expression zahlreicher synaptischer Proteine diskutiert werden. Die veränderte Morphologie der Neuronen in der lateralen Amygdala kann eine Ursache für das zwanghafte Verhalten der Mäuse sein, jedoch ist anzunehmen, dass Zwangsstörungen nicht bloß eine monokausale Ursache haben. Diese Arbeit identifiziert SPRED2 als neuen Regulator der Morphologie und Aktivität von Synapsen und die Amygdala als wichtige Hirnregion bei der Entstehung von Zwangsstörungen. SPRED2 ist somit ein vielversprechender Angriffspunkt für andere und spezifischere Untersuchungen der Hirnfunktion und eine potenzielle genetische Ursache für weitere neurologische Erkrankungen
In this present dissertation, the consequences of SPRED2-deficiency in a knockout mouse model have been investigated. In particular, the possible connection to the obsessive-compulsive disorder was examined. The SPRED2 protein is found in many tissues in the human body. Especially in the brain, ubiquitous expression was found and a connection to neurogenesis and neuronal differentiation was suspected. Its regulatory function is an inhibitory effect to the BDNF/TrkB-ERK signaling pathway, which amongst others is responsible for the transcription of neuronal genes. The SPRED2-deficient mice used were generated by insertion of a gene trap vector into the Spred2 gene. The insertion ultimately prevents the correct translation of the protein. From the SPRED2 knockout mouse line only male animals were used. As part of a SPRED2-KO study by AG Schuh of the Physiological Institute of the University of Würzburg, which showed, inter alia, the derailment of HHNA resulting in increased stress hormone levels and a dysregulation of the mineral household hormone aldosterone, obsessive behaviors were observed in the experimental animals. Subsequently, electrophysiological measurements were performed indicating an abnormality in synaptic transmission between thalamus and amygdala. Increased efficiency and excitability of the amygdaloid neurons led to the morphological investigation, which were accomplished in the context of this work. Since the afferents of the thalamus predominantly project into the lateral nucleus of the amygdala, it was first considered. The aim of the study was to find out if the knockout of the SPRED2 protein in mice leads to an altered morphology of neurons of the lateral amygdala. If so, it could at least somewhat explain the compulsive behavior of SPRED2-deficient mice. The brains of the test animals were cut into 150 μm slices and, after Golgi-Cox impregnation according to Glaser and Van der Loos, embedded in celloidin and then analyzed using a bright field microscope and the Neurolucida system. Quantitatively, pyramidal class 1 neurons of the lateral amygdala were recorded and analyzed, including the absolute number and density of the spines at their dendrites. The study showed a significant increase in the mean length of the apical dendrites in branch order 3 in SPRED2-KO mice and a tendency to increase the total number of spines on the dendrites in branch order 1-3 compared to the wild-type mice. It can be concluded that a knockout of the SPRED2 protein affects the morphology of the neurons of the lateral amygdala. The increased mean length of the apical dendrites in branch order 3 and the tendency to increased spine counts correlate with the increased synaptic transmission and excitability of amygdaloid pyramidal neurons. At the molecular level, the hyperactivity of the lateral amygdala may be discussed as a consequence of the lack of inhibition of the BDNF/TrkB-ERK pathway and the resulting altered expression of numerous synaptic proteins. The altered morphology of the neurons in the lateral amygdala may be a cause of the compulsive behavior of the mice, but it can be assumed that obsessive-compulsive disorder does not merely have a monocausal cause. This work identifies SPRED2 as a new regulator of morphology and activity of synapses and the amygdala as an important brain region in the development of obsessive-compulsive disorder. SPRED2 is thus a promising target for other and more specific studies of brain function and a potential genetic cause for other neurological disorders

Phytophthora lateralis is the causal agent of cedar root rot, a fatal forest pathogen whose principal host is Chamaecyparis lawsoniana (Port-Orford-cedar), a predominantly riparian-restricted endemic tree species of ecological, economical, and cultural importance to coastal Oregon and California. Local scale distribution of P. lateralis is thought to be associated with timber harvest and road-building disturbances. However, knowledge of the landscape-scale factors that contribute to successful invasions of P. lateralis is also important for effective land management of Port-Orford-cedar. P. lateralis is able to infest in wet conditions via stream networks (zoospore) and dry conditions via road networks (resting spore). This study tested the hypothesis that vehicles spread P. lateralis by relating its distribution to traffic intensive, anthropogenic disturbances (i.e. a road network, timber harvest) over a 31-yr period in a 3,910-km² portion of the Rogue River-Siskiyou National Forest in the Siskiyou Mountains of Oregon. Indices of road disturbance (presence/absence, configuration, length, density, road-stream network connectivity) and timber harvest (presence/absence, area, density, frequency) were related to locations of infested cedar populations from a USFS survey dataset using a geographic information system (GIS). About 40% of 934 7th-field catchments were infested with the pathogen. Total road length of the study site was 5,070 km; maximum road density was 8.2 km/km2 and averaged 1.6 km/km² in roaded catchments (n = 766). Timber activities extracted 17,370 ha (2,338 cutting units) of forest across 509 catchments; 345 catchments were cut ≥ twice. Maximum harvest density was 0.92 km²/km² ([mean] = 0.04). Both road networks and timber harvest patchworks were significantly related to cedar root rot heterogeneity. Chi-squared contingency tables showed that infestation rates were 2.2 times higher in catchments with roads compared to roadless catchments and 1.4 times higher in catchments with road-stream intersections compared to those that were unconnected. Infestation was twice as likely in catchments with both harvest and road presence than road presence alone. Single-variable logistic regression showed that a one percent increase in harvest density increased infestation odds 25% and a one-unit (km/km²) increase in road density increased infestation odds 80%. Road and stream network configuration was also important to pathogen distribution: 1) uninfested catchments are most likely to be spatially removed from infested, roaded catchments, 2) only 11% of 287 roaded catchments downstream of infested, roaded catchments were uninfested, and 3) only 12% of 319 catchments downstream of infested catchments were uninfested. Road networks and timber harvest patchworks appear to reduce landscape heterogeneity by providing up-catchment and down-catchment access to host populations by linking pathogenic materials to the stream network. Timber harvest data suggest that while infestation risk to Port-Orford-cedar populations remains high, management policies may have curbed infestation risk in timber-harvested catchments; if this is a result of specific P. lateralis mitigation policies adopted in the late 1980's or broader, region-wide conservation policies (i.e. the Northwest Forest Plan) is yet unclear.
Graduation date: 2012

碩士
國立成功大學
航空太空工程學系
103
Combustion flow control is one of the key technologies in the development of supersonic combustion ramjet engine (Scramjet). The objective this thesis is to investigate the effects of temperature and pressure changes on liquid spray behavior in supersonic flow in order to facilitate the future requirements of scramjet. The research utilizes shock tunnel to provide a Mach 2 free stream of air and JP4 is injected laterally into the free stream. The spray behaviors are compared at constant momentum flus ratio of JP4 to air stream while varying the pressure of air stream and the temperature of both air and JP4 The results show that the increase of air temperature actually increases air flow velocity so as to increase the shear stress between the air flow and fuel. Temperature also raises the viscosity of air to cause early boundary layer separation before reaching the injected liquid fuel at wall. It is observed that the interactions between the oblique shock caused by boundary separation and the bow shock caused by liquid jet are more intense and unstable, thus to induce a more unstable downstream air flow which enhances the mixing between the flow and fuel, resulting to decrease the dissipation distance and penetration height of liquid spray. Increasing fuel temperature makes liquid fuel evaporates more readily and causes flash atomization when the high temperature fuel faces sudden pressure drop while injection, and both effects decrease the dissipation distance of spray. Although part of high temperature fuel is evaporated at injection, the spray keeps most of the momentum and thus the penetration height is hardly affected. It is also observed that both the dissipation distance and penetration height decreases with the increase of air pressure. Basically, this can be comprehended to be caused by stronger impact energy of air stream to the liquid jet.