The purpose of this project is to develop a market-competitive small-scale solar-power system producing 20 W of continuous power without utilizing photovoltaic technology.  The proposed design incorporates a reflective parabolic trough and solar tube to heat a Stirling engine, which powers an electric generator attached to the load and a battery for storage.  The key technical issues include the dimensioning and material selection for the trough, engine cylinder and heat exchangers, flywheel and piston linkages, and support structure. Based on preliminary analysis, the design is feasible, but is probably not competitive with existing technology.

Product Description

The proposed product is designed to convert solar energy into 20 W at 12 V DC of continuous and storable electrical power using non-photovoltaic technology in an environment comparable to Notre Dame, IN.  The system is also designed to be low-cost, utilizing as much recycled and reused technology as possible in order encourage implementation in low-tech environments, such as in disaster relief or third-world development initiatives.  The product consists of a reflective parabolic trough solar collector to focus radiation on an evacuated solar tube, which is interfaced with an in-line beta Stirling engine.  The engine is coupled with a DC electric generator that simultaneously sources power to the load for use and sends excess power to a battery.  The entire assembly is supported by a triangular truss structure. 

The operation (see Operation page of Project Report for more detail) of the product can be broken down into three subsystems: (1) collection of the solar energy, (2) conversion of the solar energy into mechanical power used to run a generator to create electrical power, and (3) distribution of the electrical power to the load and the battery. 

• The collection subsystem is implemented by mounting onto the truss structure a parabolic trough that focuses incident light on an evacuated solar tube, which contains a heat pipe.
• The conversion subsystem is comprised of a heat-exchanger interface built into the Stirling engine that transfers heat from the bulb of the heat pipe to the hot end of the engine, which powers a generator to convert the mechanical power to electricity.
• The distribution subsystem supplies the electricity to the load and directs excess power to the battery for storage.

The key features of the design include the support structure, the parabolic trough and collector tube subsystem, and the Stirling engine.  Trade studies (also documented in the Project Report) were used to analyze the technical issues associated with each of these features. 

• The support structure is a key feature of the design because it provides the integration and interfacing of all other components.  It not only houses the trough, which must be positioned at a particular angle, but also the engine, flywheel, and generator.  After finalizing the general conceptual design of the truss support, a trade study was used to determine appropriate dimensions of the members and the material to use in order to handle a number of loading conditions. 
• The trough and tube subsystem is a key feature of the design because it is responsible for collecting the solar energy, which it can only do effectively with proper trough dimensions, reflectivity, and orientation.  The trade study associated with the trough determined optimum dimensions and materials with emphasis on the sensitivity of the solar collection to misalignment. 

The engine itself provided many technical challenges, and was consequently broken into a number of elements: the heat exchangers, the piston-cylinder arrangement, and the linkages from the power piston to the flywheel.  The challenge of designing a heat-exchanger interface between the hot bulb of the solar tube and the Stirling engine was settled through sketches and a comparison of the pros and cons associated with multiple design options, and a trade study was not done on this design issue because it was primarily conceptual. 

• The cold heat-exchanger, or heat sink, on the other hand, dealt with the effective dissipation of heat from the engine to the atmosphere, and a trade study was conducted to determine the optimum number of fins and their size and spacing.
• The study of the inside of the engine, or the piston-cylinder arrangement, concerned itself with the most efficient transfer of power from the hot bulb of the collector tube to the power piston (and ultimately to electricity via the generator).  The trade study determined appropriate dimensions for the cylinder, mass of air within, and the rpm at which the engine should run. 
• The final trade study dealt with the engine’s linkages and optimized the dimensions of them in order to obtain the desired stroke length and engine speed as determined by the piston-cylinder study. 

Only after studying these important aspects of the design was it possible to move past the conceptual stage into the actual technical details in order to create a virtual prototype in CAD and to construct a physical prototype to test the design’s feasibility.

Design Relevance

Small-scale solar power, particularly the product design investigated in this study, is relevant to any homeowner, especially those in remote locations.  The product is particularly relevant to individuals in the third-world and developing nations and disaster victims.  Solar power effectively addresses the energy needs of this demographic because it is capable of operating independent of an electric grid and requires no fuel.  Perhaps most important for the intended customers is that the cost of solar power is only the capital cost of installing the device and any maintenance associated with it.  The small-scale aspect is important because reducing the scale reduces the capital cost and it also allows for mobility of the product due to its compactness and allows for implementation based on individual needs.  The relevance of the product is, however, inherently limited by its scale of 20 W of power production, which is useful for powering light-bulbs and small appliances such as a radio and for recharging batteries.  Nevertheless, this application does meet a practical need, as evidenced by the existing products that are meant to address it.

Table 1 gives specifications for some of the prominent competition to the proposed product.  A battery would meet most needs because it is compact, requires no fuel, and fulfills the power requirement.  However, the battery is heavier and more expensive than desired.  A gas-power generator well exceeds the power requirement, and is also relatively compact, but is also more expensive than desired and most importantly, runs on gas, which incurs an additional cost and assumes availability of the fuel, which is not always the case in the third world or a disaster situation.  The human power generator, a Windstream Power product, creates electricity from human powering of bike pedals.  The product requires a great amount of physical exertion to achieve the desired power and it would be unreasonable to attempt to operate the product continuously.  Furthermore, the product is much too expensive for the intended customers.  The existing competition which most closely fulfills customer needs is solar panel technology.  The technology is lightweight, compact, meets energy requirements and comes at a reasonable cost.  Furthermore, the power it produces can be easily stored in a battery.  The relevance of the proposed design, therefore, will essentially be a measure of how well it competes with solar panel technology in these key areas.

Table 1: Benchmark Competitors for Compact Solar Power

Design Process
A key aspect of the design process involved breaking the process into manageable pieces and ordering them in a schedule of deliverables, in addition to managing the group’s resources, which included the funds provided and the time of each group member.  The overall schedule and key deliverables were determined by management, while certain important details were determined by the group. This produced the following outline of the design process:

1. Determine project goals and design requirements
2. Break design into distinct subsystems if possible
3. Explore all options for each of these subsystems and propose initial concepts
4. Analyze concepts and select final concept
5. Identify key technical issues and conduct trade studies
6. Finalize design and produce virtual model
7. Determine prototype performance requirements
8. Order parts for construction of prototype
9. Prototype product to assess feasibility

Step 1 consisted of dissecting the problem statement to arrive at the design requirements for the product (documented in Project Management under Deliverables).  As mentioned in the design requirements document, the group arrived at the presented information through intuition and web-based research.  Key aspects of the design requirements were that the product must meet certain electrical specifications (20 W at 12 V DC continuous electrical power) and standards for ease of assembly and maintenance.  Management set a deadline of 22 January for the design requirements, and the group set a deadline of 17 January for their completion.

Step 2 was crucial in simplifying the problem statement for the project.  By breaking the design into the three major subsystems mentioned in the Product Description, the group could focus on each individually.  After determining the design requirements, it became obvious that the product would be required to collect solar energy, convert it into electrical energy, and store any excess energy; thus these three requirements became the three major subsystems of the product.  This problem simplification was arrived at by the group on 17 January.

Step 3 involved brainstorming as a group the possibilities for each of the subsystems and researching individually to develop initial concepts.  For the collector subsystem, it was determined that the energy could be collected most effectively with either a parabolic trough or a parabolic dish.  For the conversion subsystem, it was determined that the solar energy could be converted to electricity through three non-photovoltaic methods: thermoelectric, thermoacoustic, and thermomechanical.  For the storage subsystem, it was quickly decided that for electrical energy a battery is the simplest and most effective method, such that the design of this subsystem was effectively complete.  After acknowledging this, the group decided that the design and especially the prototype would focus primarily on the first two subsystems and their effective interfacing.  The group set a deadline of 22 January for presenting individual initial concepts to the group.  The individual concept memos were due to management on 24 January.

Step 4 involved analyzing the pros and cons of each of the initial concepts and combining the strengths of each in order to select the final design concept.  Management set a deadline of 12 February for the design selection, while the group goal was 1 February.

Step 5 was to identify key technical challenges associated with that concept and use trade studies to assess the feasibility of those aspects of the design.  The key features, as mentioned above in the Product Description, were the support, the trough and tube subsystem, and the heat exchangers, the piston-cylinder arrangement, and the linkages of the engine.  Trade studies were proposed to management by 12 February and were conducted for each of these features with a deadline from management of 28 February and a group deadline of 21 February.  Each trade study was done separately by the individual group members to distribute the work appropriately.

Step 6, finalizing the design and creating a virtual model, was made possible by completing the trade studies, as they finalized most of the remaining details of the design, such as material selection and part dimensioning.  Creating a CAD model allowed the group to more completely conceptualize the interfacing of each of the parts and to arrive at a plan for constructing the prototype.  The group deadline for the CAD models of the components was 28 February and the CAD model of the assembly was 3 April.  The component models were done individually right after completing the trade studies, while the assembly was created just prior to purchasing parts.

Step 7 was to determine the prototype performance requirements (documented in the Deliverables) – a rubric for grading the performance of the prototype.  It was determined that the key features of the prototype would be demonstrating the collection system and its interface to the Stirling engine.  Management gave a deadline of 18 March; the group’s deadline was 28 February.

Step 8, ordering parts for the prototype, occurred on different dates due to prototype design decisions where the prototype differed from the designed product.  Parts were ordered as soon as it was determined that they would be used.  For example, the group determined early that an evacuated solar tube would be used for the collector system, and it was ordered 13 February.  The Stirling engine for the prototype was ordered on 28 February and the rest of the parts (mostly connectors and structure material) were picked up from a local hardware store on 16 March.  The group budget and bill of materials are documented under Project Management.

Step 9 was to construct the prototype and use it to assess the feasibility of the design.  The results of the feasibility analysis are documented later in this report.  The group goal for the completion of this step was 8 April, with a deadline from management of 15 April.

Options Considered
As mentioned previously, the design was decomposed into three major subsystems, such that the selection of a design concept was simplified by essentially considering each subsystem separately (this was not always possible, but the decomposition of the problem in this way was very beneficial) and matching the best collection subsystem to the best conversion subsystem and storage subsystem. 

As mentioned in the Design Process section, the battery was immediately evident as the best option for the electrical storage, and no other options were considered at length. 

For the collection subsystem, three possibilities were explored: direct collection of incident light (no focusing), focusing light via parabolic trough, and focusing via parabolic dish.  It quickly became apparent that direct collection of light would not provide enough power, so the two focusing techniques were the only seriously considered options for the collection subsystem. 

For the conversion subsystem, three techniques for converting the thermal energy output from the collection subsystem into electrical energy were explored: thermoelectric, thermoacoustic, and thermomechanical.  The thermoelectric method works by taking advantage of the thermoelectric effect, which is the production of current across two dissimilar metals subjected to a temperature gradient.  The thermoacoustic method (Figure 1) works by virtue of the fact that a temperature gradient across a regenerator (small mass with high thermal capacity) in a resonator (such as an aluminum tube) sets up a traveling acoustic wave, whose energy can be harnessed to produce electricity.  The thermomechanical method is the most familiar of the three; it is essentially a heat engine.  Within this realm three options were considered: steam engine (Figure 2), steam turbine, and Stirling engine (Figure 3).

The combination of the three major components of the system is shown in a simple schematic (Figure 4).  This schematic was drawn after choosing the parabolic trough for the collector, but before making a final engine selection.  It shows how each subsystem would be integrated into the design.  A number of other sketches and more thorough explanations of them are documented in the Initial Concept Memos under Deliverables.

Table 2 lists the positive and negative aspects of each of the options considered, highlighting key technical challenges associated with each.  The listing of these aspects was essential in selecting the final concept.

Table 2: Concept Selection Study Comparison

Solutions Selected
Based on the comparison of the relative strengths and weaknesses of the presented options, decisions were made to finalize the conceptual design, which consists of a parabolic-trough-enhanced solar tube linked to a Stirling engine.  For the collector subsystem, the parabolic trough was selected due to the requirement for only one degree of tracking as opposed to two; this greatly simplifies the design, and the fact that it only needs to be set once a day eliminates the need for diverting power to an automated tracking system.  In order that the trough will be able to concentrate heat at a point, a naturally convecting heat tube was selected to place at the focus of the parabola.  A potential weakness of the heat tube is that it operates most effectively at an incline of 70 degrees, but in order for the trough to collect the maximum amount of light, the trough (and thus the tube) must be inclined at the smallest angle for which the tube will still function, which is 30 degrees.  Another key technical issue is the sensitivity of the collector to misalignment of the trough with respect to the path of the sun.  This issue was examined in the Parabolic Trough Solar Collector trade study, which shows that a misalignment of over 4 degrees can result in complete power loss.  This implies that careful attention must be paid to proper manufacturing of the endplates and application of the reflective surface to the trough.

For the power conversion subsystem, the Stirling engine was chosen due to its likelihood of feasibility, amenability to use of recycled parts, and ability to run on lower temperature differences.  Key technical issues associated with the engine are the dimensioning of the cylinder, the engine speed and associated linkages, and the heat sink.  The Stirling Engine trade study indicated that the engine would in fact be capable of delivering the required power due to its ability to achieve an efficiency as high as 26%.  The Flywheel and Engine Links trade study indicated that initial torque to be overcome may present difficulties for engine startup, since the flywheel will weigh 41.95 lbs.  It may be practical to implement an intelligent system to kickstart the engine when appropriate.  The Annular Heat Sink study indicated that the heat sink geometry will be dependent on manufacturing capabilities, which may present certain weaknesses in that the heat sink may not be able to ensure that the temperature difference between the cold chamber of the engine and ambient air is less than 10 degrees C.  Overall, the trade studies indicated that the chosen design will be feasible.

Implementation and Feasibility

To demonstrate and test the collection aspect of the concept, a mirrored, parabolic trough was constructed at roughly half the designed scale.  A half-scale, that is, twenty-four inch, solar tube was obtained from Mr. Cary Steinkraus, an Apricus Solar dealer, and placed in the trough at a thirty degree angle at the focus of the parabola.  The mirrored surface was created using sample reflective material from Southwall Technologies and ReflecTech (material with pressure sensitive adhesive), which was glued to an aluminum flashing backing.  The parabolic ends were made of plywood and cut using a CNC mill; the frame was constructed of PVC piping.  This aspect of our prototype was tested by using it to heat a known quantity of water.  The temperature rise was monitored and timed, resulting in power values which were then divided by the incident area of the trough to obtain power per area results, summarized in Table 3.

Table 3: Solar Collector Test Results

NASA data indicates the average solar power per square meter in South Bend in April is 225 W/m2, thus the average efficiency of the trough was approximately 78%.  This test confirmed that a necessary amount of energy can be delivered through the collection system to the engine using a reflective parabolic trough and solar tube.  The efficiency could be improved by ensuring the reflective surface is completely smooth, thus focusing all incident light onto the tube.

Due to monetary, time and skill constraints, it was not possible to prototype the designed engine.  Instead, to demonstrate the Stirling cycle, a model was purchased from LittleMachineShop.com.  This engine runs the same cycle as the concept, but in a different configuration.  Despite much troubleshooting, the engine ultimately failed to run for more than a few minutes at a time.  While unfortunate that the engine could not run longer, it did confirm that in a better designed engine, the Stirling cycle is feasible to generate power.  It also demonstrated just how tight the tolerances are on Stirling devices and the great care necessary in their fabrication and material choices.  The engine also provided a platform to demonstrate part of our electronic system.

The electronics portion of the design called for a power management circuit which would direct power generated from the engine to the load and take excess power to batteries and an engine starter.  Since the designed engine was not able to be built, and the model engine did not generate any power, a power management system would not demonstrate anything.  The model engine though, did allow for the starter to be prototyped.  This device monitors the temperature difference between the hot and cold chambers of the engine, and monitors the engine itself using a photo gate on the flywheel.  If the temperature difference is high enough, the engine should be running, thus breaking the photo gate.  However, due to the high inertia of the system, it is unlikely that the engine will start on its own.  Therefore, if the temperature difference is high enough, but the photo gate is not breaking, the micro-controller will run the generator backwards as a motor, thus kick-starting the engine.  This aspect of the electronic system worked very well in the prototype with an LED indicating when the generator would be run backwards.

Conclusions

After construction of the collection apparatus, it is clear that a sufficient amount of power can be transferred from the sun into a Stirling engine using the parabolic trough set up.  Based on the trade studies completed, it is clear as well that a Stirling engine with this energy input can produce enough mechanical energy to drive a generator to produce at least 20 watts of electrical power.  Thus, this solution is conceptually feasible from an engineering standpoint.  There are, however, some specific issues that, as a result of continuing analysis and discussions, which would be changed in a final design.

First, the engine link system designed would likely bind up due to high moments developed.  As a result, a different engine configuration, such as a rhombic drive, would be used in place of the basic crank-slider mechanism analyzed in trade studies.  Secondly, while the collection system designed works, requires no power and easily tracks the sun, it would be ultimately better to have a dish-based system with active, two-axis tracking.  While this would cost power, energy would be focused to a point, allowing for a higher power per area value and delivering that power at a higher temperature.  In addition, the elimination of the solar heat tube would simplify the system and eliminate the need to have the collector angled to the horizon.  These things would offset the power required to track the sun.

In conclusion, then, the design presented here is feasible, but ultimately not competitive.  The high manufacturing costs, space requirements and system complexity are not justified by the power generated.  For these low power applications, photo-voltaic solar panels are currently the best solution.  Stirling technology, though, is still viable for large-power applications, and as research and development continues, may someday be feasible for low-power applications.