No description available.

PhD, Faculty of Engineering and the Built Environment, School of Mining Engineering, University of the Witwatersrand, 2003

The camera is the cheapest and computationally real-time option for detecting or segmenting the environment for an autonomous vehicle, but it does not provide the depth information and is undoubtedly not reliable during the night, bad weather, and tunnel flash outs. The risk of an accident gets higher for autonomous cars when driven by a camera in such situations. The industry has been relying on LiDAR for the past decade to solve this problem and focus on depth information of the environment, but LiDAR also has its shortcoming. The industry methods commonly use projections methods to create a projection image and run detection and localization network for inference, but LiDAR sees obscurants in bad weather and is sensitive enough to detect snow, making it difficult for robustness in projection based methods. We propose a novel pointwise and Instance segmentation deep learning architecture for the point clouds focused on self-driving application. The model is only dependent on LiDAR data making it light invariant and overcoming the shortcoming of the camera in the perception stack. The pipeline takes advantage of both global and local/edge features of points in points clouds to generate high-level feature. We also propose Pointer-Capsnet which is an extension of CapsNet for small 3D point clouds.

Deep mixing (DM) is the modification of in situ soil to increase strength, control deformation, and reduce permeability. Multi–axis augers and mixing paddles are used to construct overlapping columns strengthened by mixing cement with in situ soils. This method has been used for excavation support to increase bearing capacity, reduce movements, prevent sliding failure, control seepage by acting as a cut–off barrier, and as a measure against base heave. DM is effectively used in excavations both in conjunction with and in substitution of traditional techniques, where it results in more economical and convenient solutions for the stability of the system and the prevention of seepage. Although DM is currently used for excavation control in numerous projects, no standard procedure has been developed and the different applications have not been evaluated. As this technique emerges as a more economical and effective alternative to traditional excavation shoring, there is a need for guidelines describing proven procedures for evaluation of design, analysis and construction. The main objective of this research is to develop a methodology to design retaining systems using deep mixing technology. The method will be evaluated using numerical analysis of one selected case history.

This work investigates the effects of anthropogenic aerosols on deep convective clouds and the associated radiative forcing in the Houston area. The Goddard Cumulus Ensemble model (GCE) coupled with a spectral-bin microphysics is employed to investigate the aerosol effects on clouds and precipitation. First, aerosol indirect effects on clouds are separately investigated under different aerosol compositions, concentrations and size distributions. Then, an updated GCE model coupled with the radiative transfer and land surface processes is employed to investigate the aerosol radiative effects on deep convective clouds. The cloud microphysical and macrophysical properties change considerably with the aerosol properties. With varying the aerosol composition from only (NH4)2SO4, (NH4)2SO4 with soluble organics, to (NH4)2SO4 with slightly soluble organics, the number of activated aerosols decreases gradually, leading to a decrease in the cloud droplet number concentration (CDNC) and an increase in the droplet size. Ice processes are more sensitive to the changes of aerosol chemical properties than the warm rain processes. The most noticeable effect of increasing aerosol number concentrations is an increase of CDNC and cloud water content but a decrease in droplet size. It is indicated that the aerosol indirect effect on deep convection is more pronounced in relatively clean air than in heavily polluted air. The aerosol effects on clouds are strongly dependent on RH: the effect is very significant in humid air. Aerosol radiative effects (ARE) on clouds are very pronounced for mid-visible single-scattering albedo (SSA) of 0.85. Relative to the case without the ARE, cloud fraction and optical depth decrease by about 18% and 20%, respectively. The daytime-mean direct forcing is about 2.2 W m-2 at the TOA and -17.4 W m-2 at the surface. The semi-direct forcing is positive, about 10 and 11.2 W m-2 at the TOA and surface, respectively. Aerosol direct and semi-direct effects are very sensitive to SSA. The cloud fraction, optical depth, convective strength, and precipitation decrease with the increase of absorption, resulting from a more stable atmosphere due to enhanced surface cooling and atmospheric heating.

Recent studies have shown that, in a sufficiently deep gas-solid fluidized bed of Geldart A particles, gas streaming may occur allowing gas to bypass a large portion of the particle bed. Since this is a newly observed phenomenon in fluidized beds, there is uncertainty and lack of information about the various aspects of the streaming flow. The objective of the current project is to investigate the streaming phenomenon with a combination of experimentation and modeling. In the experimental part, pressure fluctuations as a measure of the fluidized bed hydrodynamics were used to study the influence of different parameters on the behavior of a deep fluidized bed. Pressure fluctuations have been measured at 8 axial locations from 4 to 150 cm above the gas distributor for bed depths and gas velocities ranging from 0.4 to 1.6 m and 0.04 to 0.20 m/s (equal to 10 to 50 times minimum fluidization velocity), respectively. Two particle size distributions with Sauter mean diameters of 48 µm and 84 µm and two distributor plates with differing percentage open area were also tested for each bed depth and gas velocity. Analysis of pressure fluctuations in the time and frequency domains, in combination with visual observations revealed that streaming flow emerges gradually at bed depths greater than 1 m. Increased gas velocity and fines content act to delay the onset of streaming, but can not completely eliminate it over the range of velocities examined. The two different distributor designs had no measurable effect on the streaming flow. The results of this study are provided in the first section of the present report.<p> In order to further investigate the nature of streaming flow, several cases of forced streams and jetting flows were designed and conducted, in addition to the natural streaming flow in deep beds. Results indicated that the natural streaming most closely resembles the imposed stream which not only the imposed stream, but additional gas added through the distributor. The case of jet flows with no additional gas resembles the severe streaming that might happen in very deep beds with the existence of completely non-fluidized regions. Application of supporting jets in addition to the main gas flow could enhance the fluidization quality to some extent, however, not enough to provide a normal fluidization. Wavelet analysis of the pressure fluctuations showed that in deep fluidized beds, bubbling activity with the typical dominant frequency coexist with the streaming flow, with a minor contribution. Wavelet findings suggested that the streaming flow can be considered to form by increasing the relative importance of one available stream of bubble activity with increasing bed depth. The results of this study are provided in the second section of this report. Further study of the streaming flow was undertaken with computational fluid dynamic (CFD) simulation of the deep fluidized bed. CFD simulation of fine Geldart A particles has met with challenges in the open literature and various modifications have been proposed to be able to model fluidized beds of these particles. In the present work, the commercial CFD codes FLUENT and MFIX were initially tested for the modeling of deep fluidized bed of Geldart A particles. However, simulation results did not show any sign of streaming flow in the fluidized bed. Subsequently, the commercial CFD code BARRACUDATM that has been claimed by the developers to be appropriate for this purpose, was tested. Due to the lack of data on the performance of this code, a simple case of modeling a freely bubbling fluidized bed of Geldart A particles was attempted first. For this purpose, four different simulation cases, which included three different numerical grid sizes and two drag models with a realistic particle size distribution were designed and tested. The simulated bed expansion, bubble size distribution, rise velocity and solid fraction were compared with commonly accepted correlations and experimental data from the literature. The results showed a promising predictive capability of the code without the need for modifying the drag model or other constitutive relations of the model. The third section of the report presents the simulation results of this study.<p> The BARRACUDA code was then used for simulating the deep fluidized bed of Geldart A particles. However, similar to the previous CFD codes tested, instead of streaming flow, bubbling fluidization was predicted. Therefore, a phenomenological model was developed to better understand the streaming flow. According to the model results, the stream represents a zone of much lower pressure drop compared to other parts of the bed, which can be a possible reason for the formation and stability of the streaming flow inside the fluidized bed. The model results showed that increasing the bed depth enhances the streaming flow, while increasing the gas velocity improves the uniformity of the bed and decreases the streaming severity. Streaming flow was found to be less severe for larger particle sizes. All of these trends are in conformity with the experimental results. These findings provide the content of the fourth and final section of this report.

Recent studies have shown that, in a sufficiently deep gas-solid fluidized bed of Geldart A particles, gas streaming may occur allowing gas to bypass a large portion of the particle bed. Since this is a newly observed phenomenon in fluidized beds, there is uncertainty and lack of information about the various aspects of the streaming flow. The objective of the current project is to investigate the streaming phenomenon with a combination of experimentation and modeling. In the experimental part, pressure fluctuations as a measure of the fluidized bed hydrodynamics were used to study the influence of different parameters on the behavior of a deep fluidized bed. Pressure fluctuations have been measured at 8 axial locations from 4 to 150 cm above the gas distributor for bed depths and gas velocities ranging from 0.4 to 1.6 m and 0.04 to 0.20 m/s (equal to 10 to 50 times minimum fluidization velocity), respectively. Two particle size distributions with Sauter mean diameters of 48 µm and 84 µm and two distributor plates with differing percentage open area were also tested for each bed depth and gas velocity. Analysis of pressure fluctuations in the time and frequency domains, in combination with visual observations revealed that streaming flow emerges gradually at bed depths greater than 1 m. Increased gas velocity and fines content act to delay the onset of streaming, but can not completely eliminate it over the range of velocities examined. The two different distributor designs had no measurable effect on the streaming flow. The results of this study are provided in the first section of the present report.<p> In order to further investigate the nature of streaming flow, several cases of forced streams and jetting flows were designed and conducted, in addition to the natural streaming flow in deep beds. Results indicated that the natural streaming most closely resembles the imposed stream which not only the imposed stream, but additional gas added through the distributor. The case of jet flows with no additional gas resembles the severe streaming that might happen in very deep beds with the existence of completely non-fluidized regions. Application of supporting jets in addition to the main gas flow could enhance the fluidization quality to some extent, however, not enough to provide a normal fluidization. Wavelet analysis of the pressure fluctuations showed that in deep fluidized beds, bubbling activity with the typical dominant frequency coexist with the streaming flow, with a minor contribution. Wavelet findings suggested that the streaming flow can be considered to form by increasing the relative importance of one available stream of bubble activity with increasing bed depth. The results of this study are provided in the second section of this report. Further study of the streaming flow was undertaken with computational fluid dynamic (CFD) simulation of the deep fluidized bed. CFD simulation of fine Geldart A particles has met with challenges in the open literature and various modifications have been proposed to be able to model fluidized beds of these particles. In the present work, the commercial CFD codes FLUENT and MFIX were initially tested for the modeling of deep fluidized bed of Geldart A particles. However, simulation results did not show any sign of streaming flow in the fluidized bed. Subsequently, the commercial CFD code BARRACUDATM that has been claimed by the developers to be appropriate for this purpose, was tested. Due to the lack of data on the performance of this code, a simple case of modeling a freely bubbling fluidized bed of Geldart A particles was attempted first. For this purpose, four different simulation cases, which included three different numerical grid sizes and two drag models with a realistic particle size distribution were designed and tested. The simulated bed expansion, bubble size distribution, rise velocity and solid fraction were compared with commonly accepted correlations and experimental data from the literature. The results showed a promising predictive capability of the code without the need for modifying the drag model or other constitutive relations of the model. The third section of the report presents the simulation results of this study.<p> The BARRACUDA code was then used for simulating the deep fluidized bed of Geldart A particles. However, similar to the previous CFD codes tested, instead of streaming flow, bubbling fluidization was predicted. Therefore, a phenomenological model was developed to better understand the streaming flow. According to the model results, the stream represents a zone of much lower pressure drop compared to other parts of the bed, which can be a possible reason for the formation and stability of the streaming flow inside the fluidized bed. The model results showed that increasing the bed depth enhances the streaming flow, while increasing the gas velocity improves the uniformity of the bed and decreases the streaming severity. Streaming flow was found to be less severe for larger particle sizes. All of these trends are in conformity with the experimental results. These findings provide the content of the fourth and final section of this report.

This work investigates the effects of anthropogenic aerosols on deep convective clouds and the associated radiative forcing in the Houston area. The Goddard Cumulus Ensemble model (GCE) coupled with a spectral-bin microphysics is employed to investigate the aerosol effects on clouds and precipitation. First, aerosol indirect effects on clouds are separately investigated under different aerosol compositions, concentrations and size distributions. Then, an updated GCE model coupled with the radiative transfer and land surface processes is employed to investigate the aerosol radiative effects on deep convective clouds. The cloud microphysical and macrophysical properties change considerably with the aerosol properties. With varying the aerosol composition from only (NH4)2SO4, (NH4)2SO4 with soluble organics, to (NH4)2SO4 with slightly soluble organics, the number of activated aerosols decreases gradually, leading to a decrease in the cloud droplet number concentration (CDNC) and an increase in the droplet size. Ice processes are more sensitive to the changes of aerosol chemical properties than the warm rain processes. The most noticeable effect of increasing aerosol number concentrations is an increase of CDNC and cloud water content but a decrease in droplet size. It is indicated that the aerosol indirect effect on deep convection is more pronounced in relatively clean air than in heavily polluted air. The aerosol effects on clouds are strongly dependent on RH: the effect is very significant in humid air. Aerosol radiative effects (ARE) on clouds are very pronounced for mid-visible single-scattering albedo (SSA) of 0.85. Relative to the case without the ARE, cloud fraction and optical depth decrease by about 18% and 20%, respectively. The daytime-mean direct forcing is about 2.2 W m-2 at the TOA and -17.4 W m-2 at the surface. The semi-direct forcing is positive, about 10 and 11.2 W m-2 at the TOA and surface, respectively. Aerosol direct and semi-direct effects are very sensitive to SSA. The cloud fraction, optical depth, convective strength, and precipitation decrease with the increase of absorption, resulting from a more stable atmosphere due to enhanced surface cooling and atmospheric heating.

A blowout is an uncontrolled flow of reservoir fluids into the wellbore to the surface, causing serious, sometimes catastrophic, problems in different types of petroleum engineering operations. If the formation's strength is low and the pore pressure is high, bridging can be a very effective method for blowout containment. In this method, the formation caves into the open hole or onto the casing and stops the flow of the formation's fluid, either naturally or intentionally. This method can be effective in deepwater blowouts where the formation has high pore pressure and considerable shale intervals with low strength. In this research, wellbore stability and fluid flow performance subroutines have been developed with Visual Basic for Applications (VBA) programming. By integrating the subroutines together, we made a simulation tool to predict wellbore stability during blowouts and, consequently, predict wellbore bridging during normal and blowout situations. Then we used a real case in the country of Brunei to investigate a field case of a bridged wellbore to validate the simulator. In addition to the field case, we used GMI SFIB 5.02, a wellbore stability software, to provide validation. In the final part of this research we studied the effect of water depth in bridging tendency during blowout for the deepwater Gulf of Mexico (GOM). Since we could not find any real data in this area, we used general trends and correlations related to the GOM. The results of our study showed that water depth delays the occurrences of breakout in the wellbore during blowouts (i.e. for greater depth of water, wellbore collapse occurs farther below the mudline). However, the depth in which collapse occurs is different for different maximum horizontal stress amounts.

When an engineering student attends four or five years of college to become a professional engineer one makes the assumption that they approach this learning process in such a way to gain the most knowledge possible. The purpose of this study is to measure the learning approach (deep versus surface) of first-year engineering students, test the impact of two interventions (journaling and learning strategy awareness) on increasing the deep approach to learning, and determine the relationship of the approach to learning on retention within an engineering program. The study was conducted using a quantitative self-reporting instrument to measure surface and deep learning at the beginning and end of the first and second semesters of the freshman year in an engineering program. Retention was measured as the continuous enrollment of a student in the second semester of the first-year engineering program. Results indicate that the first-year engineering students have a slightly higher level of the deep approach to learning than a surface approach to learning when they begin college. However, the results also indicate that the deep approach to learning decreased during the first semester and during the second semester of their freshman year. A student's approach to learning can be impacted by their prior knowledge, the teaching context, the institutional context or the motivation of the student. Results surrounding the learning strategies intervention also indicate that the first-year engineering students do not possess the strong learning strategies that are anticipated from students accepted into an engineering program with stringent application requirements. Finally, results indicate that a deep approach to learning appears to have a positive relationship and a surface approach to learning appears to have a negative relationship to retention in an engineering program. This study illustrates that incorporating learning theory and the use of current learning strategy measurements contributes to the understanding of a freshman engineering student's approach to learning. The understanding of the engineering student's approach to learning benefits faculty in establishing curriculum and pedagogical design. The benefit to the student is in understanding more about themselves as a learner.