This poster presentation focuses on the description of our project recently funded by the NSF S-STEM program. The project provides S-STEM scholarship support for academically-talented, financially-needy undergraduate students in two engineering departments at our university: the Department of Mechanical & Aerospace Engineering (MAE) and the Department of Civil & Environment Engineering (CEE). Described are several representative project activities, including six professional development seminars for S-STEM students, student design competition, and an S-STEM teaching seminar. The assessment results show that professional development seminars and student design competition motivate students to learn, and learn how to learn. The teaching seminar motivate instructors to think more about their own teaching and adopt active learning methods in the classroom.

The National Science Foundation (NSF) of the United States of America has established a Scholarships in Science, Technology, Engineering, and Mathematics (S-STEM) program to provide scholarships (financial aid) and increase academic success of low-income, academically talented students with demonstrated financial need who are pursuing associate, baccalaureate, or graduate degrees in STEM. This paper describes the overall framework of the NSF-funded S-STEM program at Utah State University, including program goals and student recruitment and selection. Over the past three-year project period, 31 students were awarded S-STEM scholarships. Students could renew their scholarships for up to 4 years. A detailed description of two co-curricular activities that were particularly designed and implemented for S-STEM students is provided. These activities include S-STEM industry seminars and a student design competition. The impacts of these activities are assessed through analysis of student comments and responses to questionnaire surveys. The assessment results show that both activities have a positive impact on S-STEM students. Four lessons learned from the program implementation are described to help engineering educators adopt these activities in their respective institutions.

It is shown that the smooth-wall boundary conditions specified for commonly used dissipation-based turbulence models are mathematically incorrect. It is demonstrated that when these traditional wall boundary conditions are used, the resulting formulations allow either an infinite number of solutions or no solution. Furthermore, these solutions do not enforce energy conservation and they do not properly enforce the no-slip condition at a smooth surface. This is true for all dissipation-based turbulence models, including the k-{\epsilon}, k-{\omega}, and k-{\zeta} models. Physically correct wall boundary conditions must force both k and its gradient to zero at a smooth wall. Enforcing these two boundary conditions on k is sufficient to determine a unique solution to the coupled system of differential transport equations. There is no need to impose any wall boundary condition on {\epsilon}, {\omega}, or {\zeta} at a smooth surface and it is incorrect to do so. The behavior of {\epsilon}, {\omega}, or {\zeta} approaching a smooth surface is that required to satisfy the differential equations and force both k and its gradient to zero at the wall.

Undergraduate required fluid dynamics and elective aerodynamics courses include substantial material on analysis techniques for forces acting on bodies in external flows. These methods include momentum integral analysis, and, for aerodynamic applications, lift computed using circulation and the Kutta-Joukowski theorem. The author presented in a previous FED meeting code development and preliminary classroom results for the implementation of a fully interactive, two-dimensional potential flow solver for flow over both rigid and flexible thin-airfoil (or sail) geometries. The intent of the development was to design a code that could be used as a virtual wind tunnel. The solver was developed in Fortran 90/95 with user interface and graphics routines developed using the high-level plotting library DISLIN for use on Windows-based computers. The analysis code solves the potential flow equations for single or multiple airfoils using a vortex panel method in which the vortex strength varies linearly along the panel and is continuous from one panel to the next. A variety of controls are available to adjust airfoil shapes and angles-of-attack. The user may also specify either rigid thin airfoil shapes, or flexible airfoils in which the final equilibrium shapes are determined by the pressure distribution. Available graphics include velocity vectors, pressure coefficient contours, and streamlines. Lift, axial and normal force coefficients are also output in the form of bar graphs. Several improvements have been implemented in the code, based on early student feedback, to improve its suitability for educational purposes in fluid dynamics and aerodynamics classes. These include pressure plot distributions over the airfoils, the inclusion of standard NACA 4-digit airfoil definitions, the output of velocity and pressure data about a closed contour for use in circulation and momentum integral analysis calculations, and improvements regarding compatibility for use on computers of widely varying screen resolutions. In this work to be presented, recent improvements to the code, and subsequent educational/student learning results based on a series of Qualtrics online student survey questions are presented. These survey questions query the students understanding of a) momentum integral analysis, b) circulation, c) lift calculations using the Kutta-Joukowski theorem, d) airfoil-to-airfoil fluid flow interactions, e) the necessity for attention to details when performing engineering analysis. The code may be downloaded for use by educators and students at other universities.

Experimental and simulated performance factors for a 48‐in.‐diameter butterfly valve were compared for various valve openings and flow conditions to determine the validity of using computational fluid dynamics (CFD) to predict butterfly valve performance factors such as pressure drop, hydrodynamic torque, flow coefficient, loss coefficient, and torque coefficient. Experimental data for the butterfly valve were obtained from the Utah Water Research Lab. Simulations were carried out on three‐dimensional models of the valve using general‐purpose CFD code STAR‐CCM+. Results show that for mid‐open valve positions (30–60 degrees), CFD adequately predicted butterfly valve performance factors. For lower valve‐angle cases (10–20 degrees), CFD simulations failed to reasonably predict those same values, while higher valve‐opening angles (70–90 degrees) gave mixed results. However, CFD simulations can provide engineers the ability to understand and predict valve performance, especially when laboratory testing may not be possible.

The Rotatable Buoyancy Tunnel (RoBuT) at Utah State University, built for validation measurements incorporating a high level of data completeness, is described along with the results from validation data sets for forced and mixed convection. One wall of the tunnel test section is heated while the other three are transparent for optical access. All boundary conditions, including geometry, wall temperature and inflow temperature and velocity, are measured and their uncertainties are reported. The tunnel's design is unique in that the test section can be inverted by rotating the entire facility to generate mixed convection with either buoyancy aided or buoyancy opposed flow. The RoBuT can also produce forced or natural convection, either steady or transient.Measurements for forced and buoyancy-aided mixed convection over a vertical heated plate are described. The RoBuT allows for simultaneous measurements of velocity, wall and inlet air temperature, heat flux measurements on the heated wall, and pressure drop across the test section. The fluid velocity is measured by time-averaged particle image velocimetry (PIV). The first validation case is forced convection since this flow is well understood. Both forced and mixed convection results are compared to published correlations and computational fluid dynamics (CFD) studies. The CFD is steady and 3-D using as-built measurements of the geometry. Experimental wall and inlet temperatures are used for CFD boundary conditions, as well as the inlet velocity and turbulence profiles. Three research groups perform simulations with varying levels of knowledge of the experimental results.

Nonsymmetric bifurcations, commonly used in piping systems to divert flow from one pipe to another, see considerable use in hydroelectric power applications and low‐level outlet works at dams. Under normal operation, cavitation at the bifurcation likely is not an issue; however, during emergency releases for dam safety or other extreme flows, the bifurcation must be able to safely pass the required flows without incurring damage attributable to cavitation. Although nonsymmetric bifurcations are widely used, the authors are unaware of any published data on related cavitation characteristics. The current study used computational fluid dynamics in conjunction with a physical model to predict conditions that would cause the onset of cavitation and to calculate head loss coefficients. Operating conditions are recommended that will allow bifurcations to operate within safe limits of cavitation. This study, although not exhaustive, provides previously unavailable data and can help designers and operators better understand the hydraulic performance of bifurcations.

Towed-water power generators are used in long-distance sailing to generate power for charging battery banks. A basic configuration consists of a spinning turbine (propeller) towed behind the boat and attached to an alternator or generator via a torque line. In the present work, a series of seven inch diameter, six inch pitch turbines were tested both on the water and using computational fluid dynamics techniques. Turbine rotation rates for both an alternator and generator charging a 12V, 80 AH battery were measured for the on-water tests at a nominal towing speed of 3 m/s. Turbine torque vs. rotation rate results obtained by solving the Reynolds-averaged Navier-Stokes equations were plotted against generator and alternator torque curves to predict operating rotations per minute and associated power generation. Predicted rotation rate results were in reasonable agreement with those measured on the water.

A typical undergraduate fluid dynamics course includes sections on Bernoulli’s equation, flow visualization, and lift and drag for external flows. The purpose of this work has been to develop a fully interactive, two-dimensional potential flow solver using a panel method for flow over thin-sail geometries to aid in student understanding of these topics. The user may specify either rigid sail shapes, or flexible sails in which the final equilibrium shapes are determined by the pressure distribution. The solver was developed in Fortran 90/95. The user interface and graphics routines were developed using the high-level plotting library DISLIN. A variety of controls are available to adjust sail shapes and angles-of-attack. Available graphics include velocity vectors, pressure coefficient contours, and streamlines. Lift, axial and normal force coefficients are also output in the form of bar graphs. The code was recently introduced in an undergraduate fluid dynamics class taught by the author, and is available from the author’s website for download.

Dual number automatic differentiation was applied to two computational fluid dynamics codes, one written specifically for this purpose and one “legacy” fortran code. Results for the simple case of a fully developed laminar flow in a channel validated the approach in computing derivatives with respect to both a fluid property and a geometric dimension. DNAD was also implemented into the JET fortran program which is available with a popular turbulence modeling textbook. Mean centerline velocity derivatives for a self-similar round jet with respect to all applicable turbulence model closure coefficients for k-ω and k-ε models were obtained.

Solutions obtained from a numerical method based on Prandtl's lifting-line theory, valid for multiple lifting surfaces with arbitrary sweep, are obtained for a number of rigid wing and sail geometries. The results are compared against solutions obtained using established vortex-lattice methods, and computational fluid dynamics solutions to the Euler equations. For the case of an untwisted, rectangular wing, numerical lifting-line, vortex-lattice, and Euler solutions were all in reasonable agreement. However, the numerical lifting-line method was the only method to predict the constant ratio of induced-drag coefficient to lift coefficient squared, which has been predicted from the analytic solution and confirmed by well established experimental data. Results are also presented for a forward-swept, tapered wing. Additional results are presented in terms of lift and induced-drag coefficients for an isolated mainsail, and mainsail/jib combinations with sails representative of both a standard and tall rig Catalina 27. The influence of the non-linear terms in the lifting-line solution appears minimal, with the exception of mainsail results when considering jib/mainsail combinations. (C) 2013 Elsevier Ltd. All rights reserved.

CFD calculations of experimental algae raceways under consideration for biofuel production studies were performed in an effort to assess the effectiveness of delta wing vortex formation as a means of enhancing vertical mixing. The impact of delta wings on the level of turbulence dissipation rate in these raceways was also investigated. All simulations were completed at a constant level of power input into the raceway. Velocity profiles as well as characteristic mixing times were used to analyze and quantify the vertical mixing both with and without delta wings. The velocity profiles and turbulent dissipation rate calculations suggest that delta wings are a viable method of increasing vertical motion. However, the mixing time results for the configuration of delta wings tested suggest that the additional mixing was insignificant to the raceway as a whole. Additional work on optimizing the use of delta wings is suggested.

Dual number automatic differentiation was applied to two different computational fluid dynamics codes, one written specifically for this purpose and one larger “legacy” Fortran code. Results for the simple case of a fully developed laminar flow in a channel validated the accuracy of the approach in computing derivatives with respect to both a fluid property, and a geometric dimension. DNAD was also implemented into the JET Fortran program which is available with a popular turbulence modeling textbook as companion software. Mean centerline velocity derivatives for a self-similar round jet with respect to all applicable turbulence model closure coefficients for both k–ω and k–ε models were obtained. The method proved to be accurate and relatively simple to implement into both the new and legacy Fortran codes.

CFD calculations were performed for a series of stirred, single use bioreactor vessels using both rotating reference frame and sliding mesh model approaches. Comparisons of quantities such as flow patterns, power numbers, and mixing times are presented. Calculations to predict mass transfer coefficients for a sparged 250L vessel were also performed using the rotating reference frame model. Results presented include those from a series of single, fixed bubble diameter calculations, and those which employed a population balance model consisting of 9 discrete bubble diameters.

Solutions obtained from lifting-line, vortex-lattice, and the Euler equations are presented for a series of rigid, thin wing and sail geometries. Initial calculations were performed for an untwisted, rectangular wing. For this case, lifting line theory, vortex lattice, and Euler solutions were all in reasonable agreement. However, the lifting-line theory was the only method to predict a constant ratio of induced drag coefficient to lift coefficient squared. Similar results were found for a forward-swept, tapered wing. Additional results are presented in terms of lift and drag coefficients for an isolated mainsail, and mainsail/jib combinations with sails representative of both a standard and tall rig Catalina 27. Although experimental data is lacking, overall conclusions are that the accuracy realized from lifting-line solutions is as good as or better than that obtained from vortex-lattice solutions and inviscid CFD solutions, but at a fraction of the computational cost. The linear lifting-line results compared quite well with the nonlinear lifting-line results, with the exception of the downstream mainsail when considering jib/mainsail combinations.

A one-equation near-wall turbulence model is used to predict a strongly heated, low-Mach-number gas flowing upward in a vertical tube. Intense wall heating causes significant and continuous fluid property variation and thickening of the viscous sub-layer. Consequently, a fully developed flow does not evolve. Existing literature reveals that two-equation models with various near-wall approaches generally perform poorly under such conditions. However, the one-equation model considered in this study is shown to accurately predict this flow through adjustment of a model constant related to the viscous sub-layer thickness.

The filtered Navier-Stokes equations were solved using the techniques of large eddy simulation to model flow through an aged pipe at a Reynolds number of 6800. The large eddy simulation produced Darcy-Weisbach friction factors that were 20% less than the friction factors obtained from experimental tests. Much of the error is believed to be a consequence of filtering the smallest roughness elements when meshing the pipe wall boundary and possible deficiencies in the subgrid-scale model at modeling the complex three-dimensional flow structures due to the irregular pipe boundary.

The Reynolds-averaged Navier-Stokes (RANS) equations were solved to model flow through two aged pipes at Reynolds numbers ranging from 6,700 to 31,000. Turbulence models employed include the v(2)-f, realizable k-epsilon, and k-omega models. The v(2)-f turbulence model was found to more accurately reproduce available experimental results compared to the k-epsilon and k-omega turbulence models for flows at R=13,000 and R=31,000, while the realizable k-epsilon model was most accurate at R=6,700. Much of the error is likely attributable to deficiencies in modeling complex flow structures with flow separation and wall roughness elements smaller than the grid scale.

A validation study for two CFD models of the time-varying flow through a confined bank of cylinders is presented. The facility mimics the lower plenum of a high temperature reactor and is arranged with the cylinders on equilateral triangles with pitch to diameter ratio of 1.7. Time-resolved Particle Image Velocimetry (PIV) measurement coupled with pressure measurements along the facilities walls are compared to both the Unsteady Reynolds Averaged Navier Stokes (URANS) k–ω model and the Detached Eddy Simulation (DES) models. Spatial (i.e. time-averaged bulk velocity and pressure losses and local velocity distributions) and temporal (i.e. dominant frequencies and correlations) validation parameters on both the local and global scale are used for validation. It is found the CFD models accurately predict frequencies present in the pressure along the walls next to the cylinders in the first and the last cylinder, yet predicts other dominant frequencies in the remaining cylinders that are not found in the experiment. The temporal behavior of the DES was generally far superior to that of the URANS model.