Key Words Student Paper, Nuclear Engineering

Transcription

1 Development of Educational Undergraduate Course Modules for Interactive Reactor Physics Nader Satvat, Purdue University Shanjie Xiao, Purdue University Kevin Muller, Purdue University Charlton Campbell, Purdue University Kevin Chesterfield, Purdue University Faculty Advisor: Tatjana Jevremovic, Purdue University Student Paper Abstract The nuclear renaissance in our country is an expected response to the overall global changes and dramatic shifts in country s assurance for sustainable energy supply. Our School witnesses an exponential increase in the number of undergraduate students as much as other schools in the country. To stay competitive and modern the undergraduate course curriculum requires innovations in teaching styles and teaching materials. The reactor physics course is usually very boring for the students due to its abstractive nature. Visualization of the reactor physics phenomena is one of the interesting and inspirational ways to teach reactor physics effectively. The web interface and web based interactive tools are under development for the undergraduate course on reactor physics. These interactive educational course modules assist students to learn about numerical reactor physics modeling remotely and with no additional software other than the regular web browser. To assure undergraduate students broad understanding and competence in the important area of reactor physics and to achieve high educational standard in teaching reactor physics we are developing various modules in a modern interactive environments. The web interfaces created for the undergraduate courses on reactor physics give the students opportunity to study the abstract concepts through: learning (how) to run different reactor physics numerical codes, connect the course materials on diffusion theory with the direct application through writing small diffusion codes or running the diffusion codes provided in the module, and what is very important through extensive graphical visualizations of the neutron flux (scalar and angular), reaction rates and neutron currents provided in the modules for various reactor-specific cases to learn about these phenomena. The visualization tools are incorporated into the interactive modules that produce graphical outputs per student s selection. For example, the neutronics phenomena related to reactor kinetics are difficult concepts to visualize in the classroom. Training an artificial neural network (ANN) using training sets from state-of-the-art neutronic modeling tools such as Monte Carlo or AGENT calculations would permit a real-time calculation of criticality parameters to be performed with minimal error. The results of this calculation can then be used to simulate changes in reactor core operating conditions. Such module facilitates often boring classroom teaching on reactor kinetics. We are also taking advantage of our research reactor (PUR-1) and all modeling and calculations are thus directly benchmarked. The suggested interactive modules development is found to be important because it offers the students possibility of analyzing interesting and challenging areas of study related to nuclear engineering and safety through less conventional ways, thus contributing to expanding and advancing their knowledge and understanding of the abstractive reactor physics. Key Words Student Paper, Nuclear Engineering

2 Development of educational undergraduate course modules for interactive reactor physics Introduction Nader Satvat, Shanjie Xiao, Kevin Mueller, Charlton Campbell, Kevin Chesterfield, Tatjana Jevremovic, Ph.D. Laboratory for Neutronics and Geometry Computation (NEGE), School of Nuclear Engineering, Purdue University To some extent we may categorize our sequence of three courses pertaining to reactor physics as the spiral in the undergraduate curriculum in that the concepts are revisited with increasing complexity from the sophomore NUCL200 (Introduction to Nuclear Engineering), to junior NUCL310 (Introduction to Neutron Transport) to senior/graduate NUCL510 (Reactor Physics) course. However, the syllabuses for these courses were developed many decades ago and do not reflect in full the nowadays multi-disciplinary dimension of these important nuclear engineering topics. In the years of nuclear renaissance all schools of nuclear engineering in the nation recognize a need for the modernized curriculum to strengthen disciplinary depth in students education in their preparation for the workforce. In order to increase the students technical proficiency, their analytical skills and motivation to study, we are developing the web-based course modules for interactive reactor physics learning. The presented web-based interactive tools are developed for and used in the junior NUCL310 class. Background We are spending time and resources trying to achieve student learning is it working? 1 The reactor physics topics are usually difficult for the students due to their abstractive nature. Visualization of the reactor physics phenomena is one of the interesting and inspirational ways to teach it more effectively. To motivate and assure undergraduate students broad understanding and competence in the important area of reactor physics and to achieve high educational standard in teaching reactor physics we are developing various web-based interactive learning modules. The web interfaces created for the undergraduate course on reactor physics give the students opportunity to study the abstract concepts more efficiently. They learn how to run different reactor physics numerical codes and how to connect the course materials on diffusion theory loaded with the equations with their direct application by running the diffusion codes provided in the modules. Through the extensive graphical visualizations students obtain better understanding of the major reactor concepts; the spatial plots of neutron flux (scalar and angular), reaction rates, neutron currents, control rod worth and phenomena related to reactor kinetics provide a unique graphical form of the reactor physics concepts. In developing the reactor physics interactive modules we are taking advantage of our research reactor (PUR-1) to test all of our new modules and tools. Students are familiar with the PUR-1 characteristics and operation through the nuclear lab-courses and by seeing again the characteristics of PUR-1 this time through numerical models is helpful in conceptual learning of the reactor physics concepts. Traditional teaching of the reactor physics includes in-class presentations of the equations and derivations with analytical examples. The undergraduate students hardly ever learn of numerical solutions and state-of-the-art approaches in nuclear reactor design and control that best illustrate the importance of the material presented in the course. With the fast growth of internet and its users, younger generations are well confident, knowledgeable and comfortable with using various web interfaces. These interfaces range from selecting classes for college and buying their needs from online shops to paying their bills and managing their stock portfolios online. Therefore, we rely on these students skills as a main assurance for easy and successful use of the developed web-based interactive course modules to enchase their learning of reactor physics topics.

3 NUCL310 Introduction to Neutron Transport Neutronics 3D Full Calculation (multi AGENT) Simple Monte Carlo Kinetics 2D Calculation (single AGENT) 2D Rectangle Assembly Calculation 3D Assembly Calculation (single AGENT) 2D Hexagon Assembly Reactor Calculation Assembly 1D Diffusion Calculation 2D Flux Distribution Time Dependent Matlab Point 2D Flux Kinetics Model Distribution & ANN Figure 1. NUCL310 web-based interactive learning modules Figure 1 shows our current learning modules available for NUCL310 course: Diffusion codes modules are developed for simple geometries for one and two dimensional diffusion equation solutions and for an arbitrary number of energy groups. Students first learn in class the theory of neutron transport by the diffusion method in one energy group and in multi-group approach. They also become knowledgeable of the analytical diffusion equation solutions for various simple geometries. Then, they are asked to use the web-based tool to solve diffusion equation for a number of different problems reflecting their understanding on reactor criticality conditions, boundary conditions, effect of number of energy groups, difference between the one and two dimensional approximation of the real geometry, as well as by plotting the flux profiles and reaction rates to explain the physics behind the plots shapes. AGENT code 2 is the state-of-the-art neutron transport code for two and three dimensional modeling of real reactor geometries. The AGENT code theory is not explained in the class in any detail as it is beyond the course curriculum. However, students are given enough comparative information to be able to understand and analyze the difference in results between the diffusion and the AGENT methodologies. This itself is a step toward modernization of the course content and improved outcomes of students learning by giving them the opportunity to learn beyond the traditional course topics. Students are given few examples to run AGENT and diffusion codes to analyze the level of accuracy. The extensive plots are available for visualization of the main reactor physics variables, such as scalar or angular neutron flux, neutron currents and reaction rates. Simple Monte Carlo is developed to illustrate the nature of neutron scattering and absorption in various materials. It represents a fast tool for students to visualize for example why light materials are selected as moderator materials. Kinetics 3 module allows student to visualize the effect of the control rod position change in the core and the effect on flux and power profile change due to reactivity insertion. Students first learn in the class about the control rod effect on reactor criticality and the point kinetics. Few homework assignments connect the theoretical learning with the practical applications through the web-based Kinetics interactive tool. Main Framework of the Web-Based NUCL310 Interface The web-based interface for NUCL310 (shown in Fig.1) is coded using PHP scripting language. The data is stored in MySQL database. Students are asked to register on the interface (Fig. 2). Once students register using their full name, student ID, and a username, the administrator will confirm their registration using the list provided by the course instructor. Once registrations for students are approved by the administrator, they will

4 be given access level to run the codes through the interface. At this point, they can login to their accounts, simply submit their jobs and wait for the interface to respond with the results. Figure 2 Registration form for students and required information for validation of the account Figure 3 Input deck editor for students to manipulate the description of the problem The interface provides the option of ing the results or downloading them through the web interface itself. This gives to the students the option of submitting their job and to check their s at their convenience. As

5 an example, we assume students are asked to run a simple 2D geometry with data available to them using the diffusion code module. As seen in Fig. 3, students are able to modify the input. At the same time, they can choose to submit their process (job) to the server by clicking on run. Once students click on run, the job will be stored in a table in the MySQL database. Then, a cron job will run every few minutes (adjustable by administrator). The cron job will run a script to check the jobs in the queue list and run them based on the time they have been submitted. Once the job is done, the interface will send the results to the address of the student or be available at the interface web page. Students are able to check the status table and the amount of jobs ahead of their job as shown in Fig.4. This way, they can always monitor when to expect results from their submitted jobs. People with administrative access can terminate jobs if the situation requires it. For instance, if a certain job has been taking too long execution time, the administrators can terminate that job to avoid other students waiting for a long time. Some Novel Aspects in Interactive Reactor Physics Learning The educational undergraduate course modules for interactive reactor physics are under development to enchase student understanding of the abstractive phenomena and reactor variables. We have developed novel visualization ways for various reactor physics variables that can help students to learn and understand the abstractive definitions of the even more abstractive words. Here are few representative examples: Angular neutron flux: Neutron in a reactor core is described with its position, r, energy E, time t at which it is observed, and direction of motion, Ω. The most detailed quantities describing the overall behavior of neutrons are time-dependent angular neutron density n ( r, E, t) and time-dependent angular neutron flux Φ( r, E, t) = υ n( r, E, t). The time-dependent angular neutron flux gives the most detailed average description of the status of neutrons in a given time and given volume. The angular neutron flux represents the number of neutrons passing through an area of a unit surface (1cm 2 ) about r perpendicular to Ω, with energy de about E and direction of motion Ω in solid angle d Ω at a given time 4. We used the AGENT code to extract this definition of the angular neutron flux and help the students visualize the flux change along different angular directions 5, 6. Figure 4 shows the angular flux profile in the fuel region pointing to its nearly isotropic distribution and in the moderator (water) region pointing to a preferential neutron scattering and its anisotropy. Neutron current: The angular neutron flux is used to define the angular neutron current. The neutron current density integrated over the directions of motions is defined with the following integral J ( r E, t) = J ( r, E, t) dω = ΩΦ( r, E, t), dω. Students learn that the scalar neutron flux 4π 4π and neutron current have the same units i.e. number of neutrons/cm 2 sec but is usually difficult to understand the difference - the current is a vector describing the net rate at which neutrons pass through a surface oriented in a given direction, while neutron flux describes the total rate at which neutrons pass through a unit area regardless of orientation of neutron direction of motion 4. Figure 5 shows a neutron current distribution across the plane in a simplified fuel assembly of the reactor core and the corresponding scalar flux distribution 5. Neutron scattering: This very important concept of neutron slowing down process and understanding of why light materials are used in thermal reactors as moderators is easily understood by running the simple Monte Carlo. An example of neutron scattering on hydrogen (light material and main constituent of water moderator) and on uranium-235 (heavy material and main constituent of nuclear fuel) is shown in Fig. 6 illustrating importance of light materials in slowing down neutrons based on a large difference in number of scattering events in these two different materials. Kinetics: The movie of the neutron flux change due to reactivity insertion and the time response represents a highly valuable supportive learning tool for students to better understand the correlations and connections between the point kinetics equations, effect on scalar neutron flux therefore reactor power, and time response. Figure 7 includes the snap shots of the movie that is presented in the class to illustrate the idea of reactor kinetics using the PUR-1 as an example.

6 Fuel Water Fuel Water Figure 4 Graphical representation of the angular flux Neutron current Fuel 1 Neutron current Fuel 2 Fast group Neutron scalar flux Thermal group 3 rd plane Figure 5Graphical representation of the current and scalar flux

7 Hydrogen Uranium-235 Figure 6. Simple Monte Carlo illustrating scattering of neutrons in light and heavy nuclei initial relative flux at t=0 seconds prompt critical flux Reactivity insertion of 1 beta over 10 seconds. Reactivity of 1 beta is held for 40 seconds, and then it is slowly decreased to zero over the next 30 seconds. At t=12 seconds the reactor goes prompt critical. The relative flux begins to decrease at t=20 seconds to thermal feedback. The relative flux remains steady until the reactivity begins to decrease, at which point the relative flux begins an exponential decay. Figure 7. What is reactor kinetics Example of the PUR-1 kinetics Conclusions Educational undergraduate web-based modules for interactive reactor physics learning were developed and implemented in the junior course. With this we offer to the students a number of new approaches in learning abstractive reactor physics concepts through their extensive graphical representations. The presented examples of the web-based interactive tools are helping modernize the course content by exposing students to computational tools that are direct or indirect application of the materials traditionally presented in the class. We intend to continue this development by including more examples from the course curriculum and by creating novel visual representations of reactor physics concepts. Especially our focus will be in extending the Kinetics and simple Monte Carlo modules, and by adding new modules related to thermal-hydraulics aspects in relation to reactor physics and reactor design, and radiation shielding aspects. Acknowledgment Research partially supported through Grant Number 2402-PU-DOE-4423 under the Innovations in Nuclear Infrastructure and Education (INIE) Program of the US Department of Energy. Bibliography M. Huirsine, S. Xiao and T. Jevremovic, Synergism of the Method of Characteristics, R-functions and Diffusion Solution for Accurate Representation of 3D Neutron Interactions in Research Reactors using the AGENT Code System, Ann of Nucl Energy, Vol 33, Pg , 2006

8 3. Mueller, K. and Jevremovic, T. Neural Network & Point Kinetics Algorithm for Virtual PUR-1 Research Reactor. American Nuclear Society Summer Session, Anaheim, CA., June 8-12, 2008, Accepted. 4. Jevremovic, T. Nuclear Principles in Engineering, Springer, 2005, Xiao, S. and Jevremovic, T. Visualization Tools for 2D/3D Full Core Neutronics Based on the AGENT Methodology. American Nuclear Society Student Conference, Texas A&M, College Station, Febr 29 March 2, Campbell, C. Xiao, S. Satvat, N. and Jevremovic, T. Angular Flux Learning Tool: Definition, Spatial Visualization and Effect of Anisotropy. American Nuclear Society Student Conference, Texas A&M, College Station, Febr 29 March 2, 2008 Nader Satvat, doctoral student at Purdue University.; designed and developed the framework for the NUCL310 learning tool, webmaster. Shanjie Xiao, doctoral student at Purdue University; developed visualization tools for AGENT and diffusion codes. Kevin Mueller, master student at Purdue University; developed the Kinetics module and associated visualization tools. Charlton Campbell, undergraduate student at Purdue University; developed the algorithm for angular flux visualizations using the AGENT code. Kevin Chesterfield, undergraduate student at Purdue University; developed the simple Monte Carlo module for neutron scattering analysis. Tatjana Jevremovic, Ph.D. an associate professor of nuclear engineering, director Laboratory for Neutronics and Geometry Computation (NEGE), adjunct professor in the Division for Environmental and Ecological Engineering, associate professor of health sciences (by courtesy), at Purdue University. Her interests and research are in advanced computational neutron transport, nuclear medicine, radiation shielding in space, environmental engineering, and ethics in engineering, global engineering and reform of engineering education.