原文(1)

Solar TrackerDavid Crowe, Jeff McCormick, Joel Mitchell,Thomas Stratton, Jeff SchwaneDecember 15, 2005Duke University Smart HousePratt School of EngineeringAbstractThe Solar Tracker team was formed in the fall of 2005 from five students in an ME design team, and a Smart House liaison. We continued the work of a previous solar tracker group. The task was to design a prototype tracking device to align solar panels optimally to the sun as it moves over the course of the day. The implementation of such a system dramatically increases the efficiency of solar panels used to power the Smart House. This report examines the process of designing and constructing the prototype, the experiences and problems encountered, and suggestions for continuing the project.1.IntroductionSolar tracking is the process of varying the angle of solar panels and collectors to take advantage of the full amount of the sun’s energy. This is done by rotating panels to be perpendicular to the sun’s angle of incidence. Initial tests in industry suggest that this process can increase the efficiency of a solar power system by up to 50%. Given those gains, it is an attractive way to enhance an existing solar power system. The goal is to build a rig that will accomplish the solar tracking and realize the maximum increase in efficiency. The ultimate goal is that the project will be cost effective – that is, the gains received by increased efficiency will more than offset the one time cost of developing the rig over time. In addition to the functional goals, the Smart House set forth the other following goals for our project: it must not draw external power (self-sustaining), it must be aesthetically pleasing, and it must be weatherproof.The design of our solar tracker consists of three components: the frame, the sensor, and the drive system. Each was carefully reviewed and tested, instituting changes and improvements along the design process. The frame for the tracker is an aluminum prismatic frame supplied by the previous solar tracking group. It utilizes an ‘A-frame’ design with the rotating axle in the middle. Attached to the bottom of this square channel axle is the platform which will house the main solar collectingpanels. The frame itself is at an angle to direct the panels toward the sun (along with the inclination of the roof). Its rotation tracks the sun from east to west during the day.The sensor design for the system uses two small solar panels that lie on the same plane as the collecting panels. These sensor panels have mirrors vertically attached between them so that, unless the mirror faces do not receive any sun, they are shading one of the panels, while the other is receiving full sunlight. Our sensor relies on this difference in light, which results in a large impedance difference across the panels, to drive the motor in the proper direction until again, the mirrors are not seeing any sunlight, at which point both solar panels on the sensor receive equal sunlight and no power difference is seen.After evaluation of the previous direct drive system for the tracker, we designed a belt system that would be easier to maintain in the case of a failure. On one end of the frame is a motor that has the drive pulley attached to its output shaft. The motor rotates the drive belt which then rotates the pulley on the axle. This system is simple and easily disassembled. It is easy tointerchange motors as needed for further testing and also allows for optimization of the final gear ratio for response of the tracker.As with any design process there were several setbacks to our progress. The first and foremost was inclement weather which denied us of valuable testing time. Despite the setbacks, we believe this design and prototype to be a very valuable proof-of-principle. During our testing we have eliminated many of the repetitive problems with the motor and wiring so that future work on the project will go more smoothly. We also have achieved our goal of tracking the sun in a ‘hands-off’ demo. We were able to have the tracker rotate under its own power to the angle of the sun and stop without any assistance. This was the main goal set forth to us by the Smart House so we believe our sensed motion prototype for solar tracking will be the foundation as they move forward in the future development and implementation of this technology to the house.2. Defining the ProblemThe project was to complete the “REV 2” design phase of the solar tracker to be used on the Smart House. While the team was comprised of members from the ME160 senior design course, the customer for this project was to be the Smart House organization. Jeff Schwane, a representative from the Smart House, was our liaisonand communicated to our group the direction Smart House leadership wished us to proceed.At our first meeting with Jeff and Tom Rose, the following needs were identified:1.Track the sun during the daye no external power source3.Weather proof4.Cost effective power gain5.Must look good6.Solar panel versatile i.e. can fit different types of panelsWith these needs in hand, we constructed a Quality Function Deployment chart. This chart can be found in Appendix A. The QFD showed the major areas of concern might have been: number of panels/size of panels, internal power requirements, motor torque required.At our first meeting we were also able to set up our goals for the semester. Having a working prototype capable of tracking the sun was to be the main goal for the end of the semester, but we soon found that in order to accomplish this, we would be forced to omit portions of the design criteria in hopes they would be worked out later. This would result in the optimization of platform space on the roof to be irrelevant, with our goal being to have one platform track. It also led to the assumption that our base would not need to be tested for stability or required to be fastened to the roof. With an idea of where we were to begin, from scratch with the possibility of using the frame from the “REV 1” design, and an idea of where we were to finish, with a moving prototype, we constructed the Gantt chart that can be found in Appendix B. Our group planned to meet with Jeff once a week to make sure we were on track with the needs of the Smart House. Jeff would also meet with Tom Rose, the director of Smart House, at least once a week in order to keep everyone on the same page. With our goals in mind we embarked on the process of idea generation.3. Concepts and Research3.1 Tracking TypeOur group used a brainstorming approach to concept generation. We thought of ideas for different solar tracking devices, which proved difficult at times due to the existing frame and concept presented to us by Smart House. Other concepts weregenerated through research of pre-existing solar tracking devices. Originally our concept generation was geared towards creating a completely new solar tracker outside of the constraints of the previous structure given to us by Smart House. This initial brainstorming generated many concepts. The first one was a uni-axial tracking system that would track the sun east to west across the sky during the course of a day and return at the end of the day. This concept presented the advantage of simplicity and presented us with the option to use materials from the previous structure (which was also intended to be a uni-axial tracker) in construction. Another more complex concept was to track the sun bi-axially which would involve tracking the sun both east to west and throughout the seasons. The advantage of this concept was a more efficient harvesting of solar energy. The third concept was to only track throughout the seasons. This would provide small efficiency gains but nowhere near the gain provided by tracking east to west.The different structures we came up with to accomplish tracking motion included a rotating center axle with attached panels, hydraulic or motorized lifts which would move the main panel in the direction of the sun, and a robotic arm which would turn to face the sun. The clear efficiency gains coupled with the simplicity of design of the uni-axial tracking system and the existence of usable parts (i.e. motor and axle) for the rotating center axle structure, led us to the choice of the East to West tracking, rotating center axle concept.3.2 StructureOnce the method of motion was chosen, it was necessary to generate concepts for the structural support of the axle. Support could be provided by the triangular prismatic structure which was attempted by the previous Smart House solar tracker group or through the use of columns which would support the axis on either side. While the prismatic structure presented the advantage of mobility and an existing frame, the columns would have provided us with ease of construction, simple geometric considerations, and ease of prospective mounting on the roof. Due to the heightened intensity of time considerations, the previous financial commitment to the prismatic structure by Smart House, and our limited budget, the presence of the pre-existing frame proved to be the most important factor in deciding on a structure. Due to these factors we decided to work within the frame which was provided to us from the previous Solar Tracker group.3.2 Tracking MotionOnce the structural support was finalized we needed to decide on a means to actualize this motion. We decided between sensed motion, which would sense the sun’s position and move to follow it, and continuous clock type motion, which would track the sun based on its pre-determined position in the sky. We chose the concept of continuous motion based on its perceived accuracy and the existence of known timing technology. During the evaluation stage, however, we realized that continuous motion would prove difficult. One reason was the inability to draw constant voltage and current from the solar panels necessary to sustain consistent motion, resulting in the necessity for sensing the rotation position to compensate. Continuous motion also required nearly constant power throughout the day, which would require a mechanism to store power. Aside from these considerations, the implementation of a timing circuit and location sensing device seemed daunting. After consulting Dr. Rhett George, we decided on a device using two panels and shading for sensed motion.4. Analysis and Embodiment4.1 Structure GeometryThe geometry of the frame was created in order to allow the solar panels to absorb light efficiently. This was done by allowing rotation in the east-west direction for tracking the sun daily and a 36° inclination (Durham’s latitude) towards the south. Because this frame was designed to be placed on a roof with a slope of 25°, the actual incline of the frame was made to be 11°.The geometry of the existing platform structure was modified. This was done in order to incorporate the results from the Clear Day Model supplied to us by Dr. Knight. This model led to the conclusion that the platform should track to up to 60° in both directions of horizontal. Thus, the angle range of the frame had to be increased. The sides of the frame were brought in to increase the allowable angle of rotation, and they were brought in proportionally to maintain the inclination angle of 11°. Also, crosspieces were moved to the inside of the frame to allow greater rotation of the platform before it came into contact with the support structure.The panels used for sensing and powering rotation were placed on the plane of the platform. Mirrors were placed perpendicular to and in between the panels to shade one and amplify the other in order to produce a difference to power the motor. The sensing panels were placed outside the platform area to maintain the largest area possible for collecting panels. A third sensing panel was mounted nearly vertical and facing east to aid rotation back towards the sun in the morning. This panel wasattached to the frame under the platform, so that during most of the day, it’s shaded with minimal effects on sensed rotation.Minimizing the torques on the motor was a main concern in order to minimize the motor power needed. The platform designed for the placement of the collecting solar panels was placed under the rotational shaft so that the panels would be aligned with it the rotational axis. Since the main panels comprise the majority of the weight putting these in the plane of the rotational axis reduces torque on the shaft. The sensing panels were placed symmetrically about the axis of rotation in order to prevent additional torque on the motor. The third panel was attached to the frame instead of the platform or rotational shaft so as to also avoid any torque.4.2 MaterialsMaterials selection for most of the frame was simple because it had already been constructed. The mirrors used for the amplification and shading of the sensing panels were also already purchased and available for use. Additional parts for attachment of the panels and mirrors to the frame were taken from the scrap pieces available in the machine shop. In our selection of sensing panels, size and power needed to be balanced effectively. The panels were to be as small as possible in order to add minimal stress and weight to the frame but also needed to be powerful enough to power the rotation of the platform. Therefore, the most powerful of the intermediate sized panels available were selected. The panels purchased also appeared to be the most reliable of our options.4.3 Drive MechanismAfter designing a prototype and testing it, the motor purchased and used by the previous solar tracker group was slipping. It was removed, and the installation of a gear system with another simple motor was suggested and attempted. Professor Knight supplied some gears as well as some belts and pulleys. One end of the shaft was lathed so that one of the pulleys could be set on it, and spacers were bought so that a 6V motor we had available could power another pulley. These pulleys were to be connected by a belt. This motor demonstrated insufficient strength to turn the rotational shaft. The original motor, once detached, was taken apart and examined. Itappeared to be working again so a new pulley was purchased to fit it and was attached in the place of the 6V motor.5. Detailed Design5.1 FrameThe frame was designed from one inch square aluminum tubing, and a five foot long, two inch square tube for the axle. It is constructed with a rigid base and triangular prismatic frame with side supporting bars that provide stability. The end of the axle is attached to a system of pulleys which are driven by the motor. It is easily transported by removing the sides of the base and folding the structure.5.2 SensorOur sensing panels are bolted to the bottom of the main solar panel frame and braced underneath with half inch L-brackets. The mirrors are attached to the inside of the sensing panels and braced by L-brackets as well. The whole structure attaches easily to the main panel frame which is attached to the main axle using four 2-inch U-bolts. A third panel is bolted to the structure to return the main panels direction towards the horizon of sunrise.5.3 How the Sensor WorksOur sensor creates movement of the motor by shading one of the panels and amplifying the other when the system is not directly facing the sun. The two sensing panels are mounted parallel to the main panels symmetrically about the center axle with two mirrors in between them. The shading on one of the panels creates high impedance, while the amplified panel powers the motor. This happens until the panels receive the same amount of sunlight and balance each other out (i.e. when the sensing panels and main panels are facing the sun.). We initially attempted using a series configuration to take advantage of the voltage difference when one of the panels was shaded (Appendix C). This difference, however, was not large enough to drive the motor. We subsequently attempted a parallel configuration which would take advantage of the impedance of the shaded panel (Appendix C) and provide the current needed to drive the motor. Once the sensing mechanism has rotated from sunrise to sunset, the third panel, which is usually shaded, uses sunlight from the sunrise of the next day to power the motor to return the panels towards the direction of the sun.6. Prototype TestingInitial testing was done using just the sensing component and a 6V motor. The panels were tilted by hand to create shading and amplification. A series configuration of the sensing panels was initially tested and proved ineffective. Data acquisition showed a maximum of a 2V difference across the motor, which was insufficient to power it. Upon testing the panels individually, it was discovered that the open voltage across each individual panel would only vary between 21.5V and 19.5V when fullyamplified and fully shaded, respectively. The current running through each panel, however, was seen to fluctuate between nearly 0 amps when shaded, up to 0.65 amps when fully amplified. Therefore, in order to take advantage of the increase in impedance of the solar panels due to shading, we chose to put our sensing panels in parallel with each other and the motor. Tests with this configuration turned the motor in one direction, stopped when the sensing panels were nearly perpendicular to the sun, and reversed direction as the panels rotated past perpendicular. We found the angle range necessary to stop the motor to be very small. It was also observed that the panels rotated to slightly past perpendicular when they ceased motion. This error may be due to a difference in the innate resistance in each individual sensing panel. When tested it was found that one panel had a resistance of 52 kΩ, and the other panel resistance was 53 kΩ. Other testing found the voltage and current provided by the sensing solar panels to the motor to be consistent at all points, excluding when the solar panels are directly facing the sun. Through testing it was concluded that resistance may need to be added to one of the panels to compensate for the differences in the internal resistances of the individual panels, and a voltage regulator needs to be added to decrease the voltage seen across the motor. The original motor was prone to failure as its slippage caused the breakdown of our initial prototype after testing. This led to the institution of the pulley and belt driven system which would allow for easier maintenance given motor failure or slippage. The success of our initial testing and prototype proved to us the efficacy of our solar tracker design.7. ConclusionThroughout this project we enlisted the support of multiple resources (i.e. ME and EE professors, previous Smart House teams). We learned early on that a clear problem definition was essential to efficient design and progress. We struggled initially as we tried to design a tracking device that was different from the previous solar tracker group’s attempt, without fully weighing the size of their investment and the advantages of using the existing frame for our purposes. As we worked with the fixed frame construction from the previous group we learned that variability of design is key, especially when in the initial phases of prototyping. After many setbacks in testing of the solar panels, we learned that when working with solar panels, much time needs to be set aside for testing due to the unpredictability of the weather.The actual implementation of using the prototype in its intended location on the Smart House roof requires weather-proofing to protect the wiring and electricalconnections from the elements, housing for the motor, a bracing system to attach the structure to the roof, and possible redesign to eliminate excess height and simplify overall geometry. The efficiency of the sensor system could be improved by widening the mirrors or by placing blinders along the sides of the panels to decrease the effects of reflected and refracted light incident on the shaded sensing panel.。