Texas A&M University-Kingsville: AKM Digital Repository
Not a member yet
1689 research outputs found
Sort by
Restoration of a retired ranch to tallgrass prairie utilizing 30 varieties of native species and the soil seed bank
No full textNative tallgrass prairies are a threatened ecosystem. The rich diversity of tallgrass prairies provides important ecosystem services that should be preserved through enhanced restoration. The first step in ecosystem restoration is the selection of locally-adapted seed sources—and this information is lacking in northeast Texas. I seeded 30 varieties of native grasses and 4 mixes of native grasses and forbs in ~500 plots on approximately one square kilometer on a retired ranch in Fannin County, Texas, in fall 2019 and in spring 2020, respectively. I compared plant density for all varieties as a monoculture and total plant density for the four mixtures. Additionally, I compared plant density between those seeded as a monoculture and in a mix. Coastal Plains Germplasm little bluestem and Lavaca Germplasm Canada wildrye established stands 2 years post-seeding at densities high enough (> 2 plants m-2) to warrant recommendation. Mixes only reached densities greater than 2 plants m-2 when seeded in the spring. I also found that almost all varieties performed similarly or better when seeded in a mixture rather than a monoculture. Historic seed banks were sampled below the ochric epipedon and below the plow layer at my study area and at similar depths at a nearby tallgrass prairie remnant. There is a historic seed bank at the retired ranch (276 seedlings m-2 ) below the ochric epipedon that is denser than that of a native undisturbed tallgrass prairie (8 seedlings m-2). Additionally, the soil seed bank is larger (mean: 326 seedlings m-2) and more variable (variance: 83456) above the plow layer than it is below (mean: 113 seedlings m-2; variance = 8180). These findings can be used to inform best practices for tallgrass prairie restoration in northeast Texas
The effect of mixxing time and energy on dispersion effectiveness and biodegradation of an oily sludge
No full textOil spills and environmental pollution from the petroleum industry have risen with the significant population increase and modernization of society, resulting in an intense need to treat oil released to the environment. Using chemical dispersants is one of the best methods to clean oil spills. Mono aromatic hydrocarbons- benzene, toluene, ethylbenzene and xylenes (BTEX) are common hazardous pollutants found in petroleum oil sludge. With the help of dispersant, mechanical mixing of oily sludge works to break down the oil sludge into tiny droplets to disperse it into the water phase, where the bacteria can degrade the hydrocarbons present. Studies have found that dispersed oil in an oil spill surface has a higher degradation rate than nondispersed oil, representing a lower environmental hazard. However, with excellent brands of dispersants present in recent times, the need to test the effectiveness of oil spill dispersants continues to grow.
This research aims to study the effect of mixing time and energy on oil dispersion and investigate biodegradation by using various types of lab-scale impellers compared to previous results
Petrophysical modelling of a braided fluvial reservoir using machine learning
No full textThe success of a predictive model depends on the understanding of the physical principles that relate the inputs to outputs. For the specific case of petrophysical properties modeling, several parameters need to be established. For example, deriving water saturation from resistivity is not so straightforward even in the simplest Archie’s model. The impact of the selection of the electrical parameters (a, m, and n), the porosity, and the water formation salinity can result in both pay and no-pay end with the same input (resistivity). This situation can become much more complicated if these parameters are not constant, in addition to the model being more robust and involving more parameters. Therefore, understanding only the input and output relationship is not enough, going directly to the origin of things helps to understand it, and here is where sedimentology works its magic.
A rock's properties are consequences of the depositional environment and its conditions such as energy, sources of sediment, and the events that happen after deposition. If the process is understood, it will be easier to establish the variability of the parameters, define them and predict the results. This thesis seeks to find the link between sedimentology and petrophysics, by denominating Petrophysical Rock types (PRTs), and incorporating them as a condition in the modeling. The methodology is proposed for a braided fluvial reservoir with a high variation of textural and mineralogical features, where it is necessary to evaluate each facies differently to obtain results that correlate with hydrocarbon production. The development of this project uses logs, core data and production data, and incorporates machine learning techniques in addition to using the models established in the oil and gas industry to improve prediction. The introduction of variations in the parameters of the models through the PRTs allows the high definition of the properties to correctly identify the potential of the reservoir
Further characterization of BMRP chimeric proteins and the residues necessary for BMRPs pro-apoptotic function
No full textApoptosis is a normal physiological process by which cells die in metazoans. Cell death by apoptosis is critical for the health of the organism, both during development and in adulthood. Functions of apoptosis include destroying old and damaged cells, eliminating immune cells that function improperly, and removing cells that have completed their role during the various stages of embryonic development. When the apoptotic process does not function properly, abnormal levels of cell death lead to diseases, including cancer and autoimmune diseases. Research to learn more about the proteins that participate in this process of cell suicide is essential for developing novel therapies for diseases characterized by abnormal levels of apoptosis, including cancer.
BMRP (Bcl-2 interacting Mitochondrial Ribosomal Protein) was discovered as a novel Bcl-2 interacting protein in yeast Two-Hybrid experiments. Through functional studies, BMRP has been shown to have pro-apoptotic activity. Studies performed with BMRP deletion mutants have determined that the N-terminal two-thirds of the protein are necessary for its cell deathinducing activity. Chimeric proteins were constructed to further delineate the region of BMRP required for its pro-apoptotic activity. These chimeric proteins fuse a portion of the human BMRP (hBMRP) protein with a region of the Drosophila melanogaster’s BMRP (dBMRP) protein. These chimeric proteins and wild-type hBMRP and dBMRP proteins have been fused to GFP for expression studies through fluorescent microscopy and Western blot analyses. Cell reduction viability assays were conducted with the chimeric proteins to measure their killing ability
Differences in web-based content for eating disorders support in division 1 and 2 institutions
No full textNationally, 20% of college students suffer from an eating disorder. Support for eating disorders on college campuses is essential to student health. The purpose of this study was to examine web-based eating disorder resources and differences between resources content for Division 1 and 2 institutions. The concept of awareness to action developed from the principles of E. St. Elmo Lewis framed the research. Addressed was the question of which categories of web-based content relating to eating disorders are present on college and university websites and if there are differences between Division 1 and 2 institutions. A mixed-methods exploratory sequential design was conducted in two phases, qualitative and then quantitative. The target
population was all institutional members of the National Collegiate Athletic Association (NCAA). Selected was a random sample of 44 Division 1 and 44 Division 2 institutions. Content analysis was the method employed for the first qualitative phase of the analysis. A chi-square test of homogeneity procedure indicated if differences existed in emergent content categories between NCAA Division 1 and 2 institutions. Of the 22 categories of web-based content tested, nine categories were observed as statistically different between NCAA Division 1 and 2
institutions rejecting the null hypothesis. 12 of the 22 categories had no difference, indicating retention of the null hypothesis. One emergent variable had no statistical outcome. The study also provided an examination and need for further research on mental health awareness,
institutional services, educational leadership, and college student well-being. The results provided contributions to the literature for the gap in web-based content for eating disorders in NCAA Division 1 and 2 institutions
A parallel algorithm for coloring the vertices of a graph
No full textA graph is an abstract data structure that consists of a set of vertices and a set of edges. Each vertex represents an entity or object, and an edge represents existence of relationship between two vertices. Vertex coloring of a graph mean assigning labels or colors to the vertices of a graph such that no two neighbors or adjacent vertices have the same color. Two vertices that are connected with an edge are neighbors of each other. Coloring the vertices of a graph using a minimal number of colors is known as the graph coloring problem. It is one of the hardest problems in the nondeterministic polynomial time complexity class (NP), also known as NP-hard. Graph
coloring problem is very important, because it is used to solve a variety of other important problems such as resource allocation or scheduling problem, circuit testing, map coloring, solving sudoku puzzles, planning seating arrangements, etc. Since this problem is NP-hard, and it takes exponential time to find the solution. Thus, for practical purposes an approximation solution is used rather than an exact solution. There are many good graph coloring algorithms available that give an approximate solution in polynomial time. But even then, for large-scale graphs, it can take a prohibitively long time to color the vertices of a graph. As a result, parallel algorithms are
necessary to color the vertices of large-scale graphs. A parallel algorithm can be very fast by distributing the computational workload among many processors. When coloring a graph in parallel using multiple processors, it is possible that in a certain iteration, neighbor vertices are
assigned the same color by different processors. This is called a conflict, and it has to be resolved. One type of parallel graph coloring algorithm first tentatively colors the vertices of a graph. Then they identify conflicts of color in adjacent vertices from different processors and resolves them. These two stages are iterated until the entire graph is colored with no conflicts. In this thesis, a parallel algorithm was developed for distributed memory systems to color the vertices of largescale graphs and reduce conflicts. This algorithm is scalable and can color the vertices of largescale graphs fast. A study was conducted on the performance of five real-world large graphs using
the algorithm developed. Various heuristics were studied to reduce the number of conflicts. Different color assignment heuristics were studied such as – initial first-fit coloring then resolving conflicts with first-fit coloring (FFit-FFit), initial random coloring then resolving conflicts with
random coloring (Rand-Rand), etc. Different vertex visiting order heuristics such as – natural order (NAT), largest degree first (LDF), incidence degree (ID), etc. were also studied. Rand-Rand-LDF heuristic combination demonstrates the least number of conflicts, but it uses comparatively a larger number of colors. The number of colors is reduced by incorporating a recoloring process called reverse coloring. When reverse coloring is combined with Rand-Rand heuristic it uses a significantly smaller number of colors than FFit-FFit. A scalability up to 160 cores is achieved using 14 processors (12 cores per processor)
Analyzing the impact of COVID-19 on the US global supply chain
No full textGlobal supply chain is vital for economic growth and development. Therefore, the need to investigate the impact of disruptions on global supply chain is evident; especially after the occurrence of the COVID-19 pandemic. Existing frameworks on supply chain risk management are not sufficient to address disruptions caused by a pandemic.
This thesis addressed the impact of Covid-19 on the US global supply chain. The Leagile Supply chain model and the Flexibility Supply chain model were evaluated. The “flexileagile” framework was proposed to mitigate the risks caused by similar disruptions in the future. Data was gathered from the United States Census Bureau database on US global trade in 2019 and 2020. Journals and peer-reviewed empirical studies were also valuable data sources. The data collected was analyzed using Microsoft excel, fish bone diagram and a decision matrix. Similarly, data gotten from Organization for Economic Cooperation and Development on surgical facemasks supply chain was used as a case study to demonstrate the application of the flexileagile framework.
The findings showed that the containment factors (lockdown and border closure) affected the global supply chain. Also, the results demonstrated the heavy dependence of the US economy on China. Similarly, Logistic & Distribution, Supply Shortage and Unavailability of Labor risks had the highest impact on the US global supply chain during the pandemic. Whereas, Price Inconsistency and Data & Information issues followed, respectively
Remote data acquisition using UAVs and Arduino systems for agricultural purposes
No full textThis thesis focuses on developing a low-cost Remote Data Acquisition System to monitor the soil parameters in an agricultural field using Unmanned Aerial Vehicles (UAV)
and Arduino systems. The sensor data acquired from various nodes are soil parameters such as soil moisture, soil temperature, soil electrical conductivity, soil pH, etc. The sensor nodes are installed across multiple locations in an agricultural field. The sensor data are stored locally on
an Arduino board powered by a solar panel located above it. The Arduino board equipped with transmission capabilities using a Wi-Fi module can send sensor data within a certain range. The sensor node is 3D printed and mounted on a pole visible to a UAV operator via First-
Person View (FPV) systems. A UAV is used to collect this data using a data receiving system consisting of an Arduino board, a data storing device, and a Wi-Fi module. The sensor data can be retrieved when the UAV returns to its home position through the SD card or the data
receiving system uploads the sensor data to an access point. The acquired data will help in the study of soil for better crop health as well as to monitor and control the amount of water used for the crops
Elliptical machines using adjustable linkages
No full textElliptical machines are exercising or training machines that are used to imitate walking, jogging, running, or climbing exercises. Different from treadmill machines, elliptical machine users never leave their feet away from the pedals, which reduces the pressures to the ankle, knee, and hip joints, and significantly decreases the impact injuries of joints. The configuration of the elliptical motion commonly mimics the natural paths of the ankle, knee, and hip joints for walking, jogging or running, which further lowers the strains and stresses on the joints. In addition to the lower joint impact, a unique feature of elliptical machines is their integrated leg and arm movements that provide full-body (dual lower and upper body) exercises. Users of elliptical machines not only exercise their legs but also push and pull the handlebars to strengthen their arms. Unlike treadmills, ellipticals are usually self-powered by user-generated motion and do not need motor and belt conveyance. They have cost and maintenance advantages over treadmills. This thesis research is on analyzing and simulating elliptical machines.
The closed elliptical trajectories of elliptical machines are generated through their linkages. The shapes of the closed trajectories depend on the linkage types and dimensions. The relationships between linkage dimensions and elliptical shapes and sizes are complicated and are difficult to represent using analytical expressions. The complexities and difficulties put challenges on designing elliptical machines. The elliptical trajectory including stride length of an elliptical machine needs to meet the requirements for different exercises, and various short and tall people with a wide range of arm or leg sizes. If an elliptical machine has fixed linkage dimensions, its elliptical trajectory has only one shape and size, which does not provide flexibility. For an elliptical machine to have flexibility, its linkage has to be adjustable. Adjustable linkages are more difficult to analyze and synthesize than linkages without adjustability. This thesis research is motivated to surmount the challenges facing elliptical machines. The research objective is to improve the performances of the current elliptical machines. In this thesis research, different types of elliptical machines (front, rear, and central arrangements) without and with adjustability will be analyzed. Their elliptical output motions will be simulated and compared. The research results from this thesis research will provide useful guidelines for developing and promoting elliptical machines
The effect of defects on traditional and 3D printed composite performance
No full textThe use of Polymer Matrix Composites (PMC) has grown immensely over the past fifty years. Additive Manufacturing (AM) (3D printing) has also flourished and can be used to manufacture PMC. Both contain defects. It is unknown if common defects (air voids, resin rich areas, delaminations) affect AM PMCs the same as traditional PMCs. This work compares the flexural/tensile response of pre-preg laminates with FDM-type AM PMC specimens with one of three defects. Two strongest chopped fiber-reinforced nylon filaments, and a graphite/epoxy unidirectional pre-preg were used to fabricate “Control/Defective” tensile/flexural specimens. An optimized print profile and drying the AM filaments significantly increased flexural (likely tensile) properties. Their absolute strengths/stiffness’ are different, some failure trends are similar. Results show common fracture locations for all tensile/flexural specimens containing air or resin rich areas. Defects decreased Toughness (strain-to-failure) in tensile/flexural tests except XStrand air voids. Interestingly, the flexural stiffness of specimens with defects increased for AM specimens, but decreased for pre-preg specimens. Ultimate strength decreased the most in the resin rich areas for both types of specimens for all materials. Pre-preg tensile tests could not be completed due to tab failures; all defects significantly reduced tensile properties for AM specimens. Type of loading, defect location, and FDM fabrication techniques can significantly affect mechanical properties