PBJ Xaustors - University of Idaho



PBJ Xaustors

Exhaust Waste Heat Recovery

ME 426, Fall 2003

[pic]

Team:

Robert Wiegers, James Stewart, Peter Jorg

Mentor:

Jeremy Boles

Client:

Frank Albrecht

Executive Summary

Government regulations stemming from environmental concerns and worldwide oil consumption are forcing the automotive industry to make improvements in tailpipe emissions and vehicle fuel economy. In a typical vehicle, the exhaust gas contains one of the largest portions of wasted energy, approximately 34% of available energy from the fuel. As a result, we have constructed a thermoelectric generator to capture this waste heat energy in the exhaust and turn it into useful electric energy.

With the thermoelectric generator mounted between the exhaust stream and the cooling channel, the generator we have built is projected to have a power output in the range of 70 to 90 watts depending on driving conditions. However, system testing requires implementation of the customized DC-DC converter currently in development. The generator is designed to be passive system, fully integrated into the Future Truck exhaust stream, with no customer action required to operate.

We have performed detailed experiments on the generator and our testing results lead us to recommend that additional funds be spent on an expanded implementation of the thermoelectric generator in order to achieve results that will positively impact the Future Truck performance at competition.

We have accomplished our objective of capturing the exhaust waste heat and turning it into useful electrical power. The long term action should be to wait until second generation thermoelectric chips are available at a reasonable cost and then continue the construction of another generator with the new chips. Until then we recommend that the thermoelectric generator be installed into the Future Truck to showcase the talents and engineering abilities of the University of Idaho mechanical engineering students.

Table of Contents

Section: Page Number:

1.0 – Background………………………………………………… 1

2.0 – Problem Definition………………………………………… 2

2.1 – Objectives……………………………………………… 2

2.2 – Constraints…………………………………………….. 2

3.0 – Concept Selection ..………………………………………… 3

3.1 – Introduction……..……………………………………. .. 3

3.2 – Absorption Cooling ……………………………………. 3

3.3 – Adsorption Cooling ………………………………….… 4

3.4 – Closed Loop Steam…………………………………….. 5

3.5 – Exhaust Flow Turbo……………………………………. 6

3.6 – Thermophotovoltaics…………………………………… 6

3.7 – Thermoelectrics…………………………………………. 7

3.8 – Decision Summary……………………………………… 8

4.0 – Product Description………………………………………… 8

4.1 – Introduction ……………………………………………. 8

4.2 – Thermoelectric Modules……….……………………...... 9

4.3 – Cooling Hardware………….…………………………… 10

4.4 – Thermal Bypass…………………………………………. 11

4.5 – Load Matching Circuit..…………………………………. 11

4.6 – Generator Section and Assembly.………………………. 12

4.7 – The User, Society, and the Environment..………………. 13

5.0 – Product Evaluation…………………………………………... 14

5.1 – Design Failure Modes and Effects Analysis ………… … 14

5.2 – Testing Results……………………….....……………… 15

6.0 – Economic Analysis………………………………………… 16

6.1 – Expenditures to Date………..…………………………. 16

6.2 – Labor Estimates.……………………………………….. 16

6.3 – Production Costs……………………………………….. 17

6.4 – Case for Implementation………………………………… 18

7.0 – Conclusions and Recommendations………………………… 19

Section: Page Number:

Appendix A – Drawing Package………………………………….. 21

Appendix B – Thermoelectric Performance Model………………. 32

Appendix C – Impact of Additional Weight Model………………. 37

Appendix D – Design Failure Mode and Effects Analysis………... 40

Appendix E – Testing …………………………………………….. 45

Appendix F – Résumés of Team Members……………………….. 49

Appendix G – Project Timeline…………………………………… 53

1.0 – Background

Environmental concerns and worldwide oil consumption are creating government regulations that are forcing the automotive industry to make improvements in tailpipe emissions and vehicle fuel economy. In contrast, consumers’ expectations in vehicle performance and capabilities are driving the energy consumption rate to ever-increasing levels. Automotive Original Equipment Manufacturers (OEMs) are now planning to incorporate telematics, collision avoidance systems, vehicle stability controls, navigation, steer-by-wire, electronic braking, additional powertrain/ body controllers, sensors and other electronic subsystems into the vehicle. Current electrical systems for mid- and large-sized vehicles currently have average electrical power loads of 0.8 to 1.5 kW and peak power loads of approximately 2 kW. In five years, the average power load is projected to be 3 to 5 kW with peak power loads of over 7 kW. Future automobiles, therefore, must either generate more power or utilize power management schemes to support safety features and accessory electronics while reducing fuel consumption and emissions.

One way of accomplishing this energy recovery in vehicles would be to incorporate thermoelectric devices into the exhaust system. In a typical vehicle, the exhaust gas contains one of the largest portions of wasted energy, approximately 34% of available energy from the fuel. When the thermoelectric generator is mounted between a heat source and a cooling channel, electrical power on the order of several watts to several kilowatts can be produced depending on the design and thermoelectric materials used in the system. To date, thermoelectric power generators have been primarily used in space and military applications. Overall efficiency of 5% has made these devices somewhat impractical and expensive for automotive mass-production application. Porsche, General Motors, Nissan, and others have previously investigated thermoelectric power generation, and have concluded similarly. Recent developments by HI-Z Technology, PACCAR Technical Center, and others, however, have changed the landscape, enabling new applications to be considered.

Continued research in development of efficient thermoelectric materials is very important if thermoelectric is to be applied in the automotive industry. An efficient thermoelectric device will make significant impact on the automobile industry. Converting what would otherwise be wasted heat energy to electrical power will make vehicles more efficient, thereby reducing dependence on foreign oil.

Sponsored by the Ford Motor Company and US Department of Energy, the annual Future Truck competition embodies the goals of increased fuel economy, reduced dependence on foreign oil, and drastic reductions in overall vehicle emissions in the light truck and SUV market. Future Truck competition is a forum for innovation and experimentation, wherein student teams take design risks outside the realm of practicality for large scale vehicle manufacturers. Therefore, thermoelectric waste heat recovery is clearly in line with the spirit of the competition.

2.0 – Problem Definition

Approximately 34% of the energy consumed by a typical internal combustion engine exits the exhaust system as waste. If a fraction of this energy could be captured for constructive use, the overall efficiency of the vehicle could be improved. As vehicle efficiency is one of the primary goals of the Future Truck competition, our challenge is to design and implement a system that will use the exhaust waste heat and create a useful form of energy that the Future Truck can use.

2.1 – Objectives

• Power generation will be as great as is feasible

o Opportunities for testing limited by lack of access to appropriate thermal sources and current unavailability of load matching circuit; projected output for current hardware is between 70 and 90 watts

• Off-the-shelf components and preexisting scientific knowledge will be utilized wherever possible

o Thermoelectrics purchased from Thermonamic Electronics Co., Ltd. of Xiamen, Fujian Province, PRC; exhaust bypass valves purchased from Quick Time Performance; exhaust supplies from JC Whitney and Burneel

• Compatibility and integration with other Future Truck subsystems will be realized

o Controls and load matching will be consolidated onto one independent board; exhaust piping is routed along passenger side of vehicle and does not interfere with hybrid or 4WD systems

• Gross vehicle weight will be minimally impacted

o Current hardware weight is 44 lbf

2.2 – Constraints

• Ground clearance and crush zone competition requirements will not be violated

o Installed system will remain above the bottom plane of and within the vehicle frame rails

• Effectiveness of catalytic converter will not be adversely impacted

o Consolidated 3-way catalytic converter will be located upstream of system, with no additional converters downstream, thus thermal alterations to the exhaust flow caused by the system will have no adverse effects

• Overall vehicle score will be positively impacted (i.e. fuel economy points exceed weight penalty)

o Impact to competition points in static events is indeterminate, as it is largely subjective; assessment of impact to dynamic event score will require further, integrated vehicle testing

• Total added weight of the system will not exceed 50 pounds

o Estimated additional weight, as distinct from total weight, is 33 lbf

• Expenditures will not exceed $2,032 (phase one budget implementation approval)

o Expenditures to date stand at $1,655.49; costs for load match circuits are yet to be realized and are estimated at approximately $250

3.0 – Concept Development

3.1 – Introduction

Work on concept development paralleled the formulation of project goals. The original mandate of the project was to recovery energy from the exhaust flow of the vehicle that would have otherwise been wasted in an effort to improve overall vehicle efficiency. Thus the original concept development was broader than the final goal statement would suggest, including as it did ideas for heat driven cooling systems in addition to power generation schemes.

The methodology of concept development centered around brainstorming with regard to the nature of the energy available and the possible means of utilizing it, followed by largely internet based research to determine the degree to which such concepts had already been developed as well as the commercial availability of necessary subsystems or components. Given the project scope and resources, the challenge was not to develop a unique process or technology but to select an appropriate technology and tailor it to the system constraints.

3.2 – Absorption Cooling

As the exhaust energy recovery was not originally limited to electrical power generation, a number of cooling schemes were investigated as a means of replacing the vehicle’s current air conditioning system, which constitutes a significant power draw. The ammonia absorption refrigeration cycle was considered as inspired by the propane-powered refrigerator. Rather than burning fuel to drive dissolved ammonia (other chemicals such as lithium bromide would also be potential options) out of the water stream, waste heat from the combustion of fuel in the vehicle’s engine would instead be employed, as is shown in figure 1.

[pic]

Figure 1. Schematic representation of automotive absorption cooler.

This type of cooling system would derive a large portion of its motive power from waste heat rather than the useful mechanical energy of the engine, as the belt-driven air conditioning unit does, however, it would also require a small amount of mechanical power to drive the pump. While the use of waste heat for refrigeration is very attractive from an exergetic viewpoint, the near constant flux of thermal conditions inherent to a typical city drive cycle makes the prospect of achieving satisfactory performance appear rather unlikely. Additionally, during conditions such as idling in a traffic jam, the system would be unable to produce sufficient cooling. This would suggest the parallel installation of standard AC unit, which would serve to further worsen what would probably be an already problematic weight issue.

3.3 – Adsorption Cooling

Adsorption cooling parallels absorption cooling except that the working fluid is adsorbed into a highly porous sorbent material rather than absorbing into a secondary fluid stream. The sorbent material used must have a very large surface area per volume to be effective, with activated charcoal and zeolites being common choices. The working fluid, as a low-pressure vapor, is adsorbed into the sorbent bed until it reaches saturation. A valve then closes the sorbent bed off from the evaporator and heat is applied to purge the working fluid from the sorbent material. After which, the high-pressure vapor moves on to the condenser. Thus, the sorbent bed is referred to as a thermal compressor. In the case of a waste heat energy recovery scheme, the heat used to purge the sorbent beds would come from the exhaust system.

[pic]

Figure 2. Flow schematic of an adsorption cooler.

Due to the discontinuous nature of the operation of the thermal compressors, shown above in figure 2, continuous operation of the overall system requires the implementation of multiple compressors in parallel. This, in combination with the relatively low coefficient of performance when compared to vapor compression cycles, necessitates that a competitive adsorption cooling system be quite large in volume. An engineering team at California State University Long Beach has conducted an in depth investigation of this concept. The physical size of their system tends to obviate the implementation of adsorption cooling on the vehicle, given the spatial constraints.

3.4 – Closed Loop Steam

As the exhaust system typically operates at temperatures well in excess of the boiling point of water, a Rankine cycle device was an obvious option. Power output from the shaft of the turbine would not likely have sufficient torque to be of direct mechanical use, however, it could be used to turn a small DC dynamo to create auxiliary electrical power for the vehicle’s subsystems.

Despite its familiarity, a steam power system did not seem likely to meet the constraints of the design space, specifically weight, volume, and the feasibility of implementing off-the-shelf technology. Two additional problem areas identified were the strong potential for maintenance concerns and the issue of variable input energy. Unlike a stationary generation facility, the heat input provided to a vehicle exhaust waste extraction steam power generator would be highly irregular and transient in nature. Developing a system to mitigate or adapt to these fluctuations appeared difficult.

[pic]

Figure 3. Textbook type figure showing a simple Rankine cycle generator.

3.5 – Exhaust Flow Turbo

The concept of using a modified turbo charger is relatively simple in nature. Standard turbo chargers use the flow of the exhaust gas through the pipes to turn a turbine that drives a compressor through which the incoming air to the engine first passes. In a power generation scheme, the turbine would drive a generator rather than a compressor. Thus, the flow of the exhaust gases would be used in a direct, mechanical method to generate electricity. This electricity could then be fed back into the 12V battery system.

[pic]

Figure 4. Off-the-shelf components for an exhaust turbo-compressor.

While the implementation of such a system appears to be relatively straightforward, with a strong potential for the employment of preexisting components, it would require a relatively large pressure differential to be effective. This exhaust backpressure on the engine would certainly require modifications to the tuning of the engine at minimum, and it presents a potential for engine performance degradation. As this represents a serious concern with regard to systems compatibility, the exhaust flow turbo concept has been discarded.

3.6 – Thermophotovoltaic Power Generation

While typical photovoltaic cells enable the generation of electricity from visible light, it is possible to devise and construct PV cells that instead operate in the infrared spectrum. This is the basic operating principle behind thermophotovoltaic power generation. A high temperature heat source is used to excite an emitter, usually a specialized ceramic, to produce infrared radiation of the desired wavelength for peak thermophotovoltaic efficiency. The emitter is surrounded by TPV chips, or the chips are surrounded by the emitter, and power is generated in much the same way as solar power plant. The primary difference is that the source is only inches away rather than millions of miles, resulting in a much higher energy flux. A team at Western Washington University has conducted extensive work on TPV implementation in vehicles. Theoretical and optimization work is being pursued in both Germany and Russia.

[pic]

Figure 5. Artist rendering and cut-away of a proposed thermophotovoltaic generator.

This technology is attractive. However, there are a number of challenges to implementation in this project. The first issue is one of availability. While research is on-going, commercial application has not yet developed, thus acquiring the TPV cells might prove to be impractical or prohibitively expensive. The second issue is more fundamental. Just as a typical PV cell will not produce a flow of electrons no matter how much red light is applied, so too will the TPV cells not function if the spectral band is not correct. Vehicle testing has proven that exhaust system temperatures under normal operating conditions are well below the required excitation temperatures. For these reasons the thermophotovoltaic power generation concept has been dropped.

3.7 – Thermoelectric Power Generation

Also referred to as thermionic devices, the function of thermoelectric devices is similar to the behavior of thermocouples, with the dissimilar metals replaced by the P and N zones of an array of semiconductors. A temperature differential across the junctions of a thermoelectric device induces a voltage potential. Early commercial interest in thermoelectrics was focused on solid-state cooling applications. When current is passed through a thermoelectric device, one side becomes hot and the other cold. Power generation is achieved by instead providing the temperature differential and allowing the current to flow into a circuit.

For their gross power output, thermoelectric generators tend to develop relatively high amperage and low voltage in comparison with more typical generation methods. The performance of the devices is strongly impacted by the temperature differential maintained across the device, with total power output roughly exponentially proportional to power output. For a given temperature differential, higher efficiencies will be achieved with a lower cold side temperature. Also of concern is the maximum sustainable operating temperature of the devices, as thermal degradation becomes problematic for currently available devices at temperatures in excess of 500 ºF. Efficiency of the power output of the devices is also affected by the electrical load against which they are working in the circuit. The optimal load is a function of material properties of the device and the temperature differential.

[pic]

Figure 6. Typical thermoelectric power module.

In an exhaust waste heat recovery system, an array thermoelectric devices would be fixed to the exterior of the exhaust piping, with some manner of heat sink implemented on the cold side to maintain as large a temperature differential as possible. The electric power thus generated would then be fed back into the vehicle’s electrical systems to relieve the demand on the alternator or to supplement hybrid energy storage systems.

3.8 – Summary of Decision

Thermoelectric power generation was ultimately selected due to its simplicity and very low maintenance requirements (solid state, no moving parts), close match to the thermal conditions inherent to the exhaust, commercial availability, and relatively small weight and volume. All of the other options, as noted, either did not match up with the expected thermal conditions, were too large or heavy, or otherwise conflicted with project constraints.

4.0 – Product Description

4.1 – Introduction

Figure 7 below shows the basic structure of the project as it interfaces with the Future Truck vehicle from an energy flow perspective. Overall vehicle performance is affected in that electrical power from the thermoelectrics to the 12 volt battery reduces alternator load, which is a component of the peripheral systems power flow. By reducing this load, the overall power draw on the engine is reduced for a given level of performance and hence fuel economy is improved. The decision was made to power the vehicle battery rather than the hybrid ultra-capacitor bank primarily because the voltage on the capacitor bank is time varying with application of electric assist. The achievable output voltage of the thermoelectrics matched much more closely with the battery than the capacitor bank.

[pic]

Figure 7. Energy flow diagram showing thermoelectric power generation.

4.2 – Thermoelectric Modules

Of critical importance to the design of the project was the selection of which thermoelectric modules to purchase. The final decision was between Hi-Z of California and Thermonamic Electronics Co. of Xiamen, PRC. As the Thermonamic modules cost approximately half as much as their competitors and budget constraints were appreciable, financial considerations played a primary role in deciding for Thermonamic. Although the gross power output of the Chinese modules was lower than that of the high end Hi-Z module, the Thermonamic modules had the advantage of a smaller size and a higher voltage to current ratio for a given amount of power. The latter consideration was important in terms of load matching circuit efficiency, where losses increase with the increased voltage differential across the DC-DC converter.

Sizing of the array of the thermoelectric modules to be employed was dictated by the thermal conditions, the electrical demands, and the financial constraints. Exhaust temperature data from the Future Truck vehicle, see figure D5 on page 48, indicated that it would be possible to expose most of the generator section to thermal conditions just below the manufacturer rated maximum continuous operating temperatures for much of the drive cycle. Given the performance modeling of the devices, see figure B1 on page 35, it seemed reasonable to assume a power output per chip of between 9 and 15 watts, with a wide range of parameters in effect. To achieve the desired 180 watt output, 16 to 20 modules would be needed, with an initial implementation of 8 due to budget constraints. As expected voltage output per module is in the 3.0 to 4.5 volt range, the modules are wired in series banks of four to bring the composite output close to 12 volts before passing through the load matching circuit to the battery. A rough schematic of this is given in figure 8 below.

Figure 8. Basic wiring schematic for thermoelectric generator.

4.3 – Cooling Hardware

With the hot side temperature a given due to the thermal behavior of the exhaust system, creating an effective means of maintaining a lower cold side temperature is a critical issue. The possibility of implementing a water jacket cooling arrangement was attractive given the much higher thermal capacity and convective heat transfer coefficients of liquids in comparison with air; however, a forced air system was selected due to the problems a liquid cooled system would run into with regard to weight, size, and budgetary constraints. It may have been possible to tap into the preexisting cooling system of the vehicle, with its large rooftop radiator, but such a possibility was dimmed by the fact that this would have required modification of other vehicle subsystems and coordination with other Future Truck team members who were not in a position to work with us early in the semester.

The cooling scheme pursued involves ducting air from the front of the vehicle to generator section located about mid-vehicle via flexible metal ducting. Air would be forced into the system by the forward motion of the vehicle, entering the ducting at two custom scoops that we fabricated by modifying the existing plastic fog light cover plates. This hardware is ready for installation; however, there is some doubt now whether routing will be possible for the driver side duct given the spatial needs of the four wheel drive hardware, which have changed a number of times as that project work has altered its focus. Another possibility is to scoop air from directly below the vehicle. This configuration, identified as a possibility early in the design, was suggested by the visiting Future Truck Ford mentor. It would require further prototyping and likely road testing, as turbulent air flow under the chassis is uncertain. To protect the heat sink from fouling and impact damaged caused by road debris, a light sheet metal box duct is used to shroud the heat sink fins. While the finalized nature of its implementation still awaits road testing and the control hardware, a 12 watt parasitic cooling fan will be installed parallel to the air intake to provide supplemental cooling.

When the thermoelectric modules were purchased, the manufacturer offered the possibility of building heat sinks to match. With the mindset of using off-the-shelf technology when available, two of these heat sinks were purchased sight unseen. These heat sinks suffered from an overly thick base plate and an extremely high fin density that all but inhibited any axial air flow. For these reasons a custom heat sink was fabricated, using a thinner aluminum base plate and a much lower fin density. Copper was selected for the fin material due to its superior thermal conductivity. The fins were attached to the base plate using a mild press fit and metal bonder, and thermal contact resistance was reduced by use of thermal grease.

4.4 – Thermal Bypass

Provisions for some means of thermal isolation from the exhaust stream were recognized as an important aspect of a survivable system early in the design. As the thermoelectric modules experience thermal degradation with prolonged exposure to temperatures very much in excess of 500 ºF, or intermittent exposure to temperatures in excess of 620 ºF, the exhaust system of the vehicle represents a dangerous environment, with the potential for all or most of the modules to be damaged or destroyed during high engine load conditions such as accelerating a loaded vehicle from a stop. A number of solutions were examined, with a bypass solution finally being selected. Air injection was considered, but the prospect of injecting enough ambient air into the exhaust flow to cool it from as high as 1100 or 1200 ºF down to around 500 ºF appeared prohibitive. A scheme for a mechanical separation of the modules from the generator section was also considered, but problems of properly reseating the modules would likely have required a separate mechanism for each chip, thus requiring a complete redesign of the heat sink. Also, problems of debris intrusion onto the module surface or thermal warping of the generator section would have proved problematic, as such occurrences could easily cause mechanical failure of the modules when a compressive load was reapplied.

A bypass arrangement allows the exhaust flow to be directed away from the generator section as thermal conditions dictate. Initially, the generator section will be open and the bypass closed, a threshold temperature will cause the bypass section to open, and finally a critical temperature will cause the generator to close. After significant research into commercially available hardware, electromechanical exhaust cut-out valves from Quick Time Performance were selected for their robust design and quality construction. During testing, a LabView control scheme that could interface with the Compact Fieldpoint systems on the Future Truck was developed to drive valve actuation based on thermocouple readings. However, recent work on the load matching circuit indicates the valve control will be integrated with this custom board.

4.5 – Load Matching Circuit

Thermoelectric devices are very sensitive to the electrical load they are subjected to in terms of optimal power generation. A match (or matched) load condition occurs when the effective electrical resistance each module “sees” is equal to that of its own internal resistance. Early research indicated that the best approach to such a circuit would be a DC-DC converter, and it was thought that given the immense array of electronic equipment available that an appropriate device could be purchased directly. Extended research into the possibility of purchasing the desired hardware was unsuccessful, with no preexisting hardware capable of the necessary function found. All was not lost, however, as recent work with electrical engineering students Jeremy Forbes and Erik Cegnar is moving towards an in-house solution to the problem. Both Mr. Forbes and Mr. Cegnar are experienced practitioners of the electrical arts, with extensive experience with the hybrid systems of the Future Truck vehicle. In return for their assistance, we will be working on mechanical installation issues pursuant to electric hybrid systems.

The load matching circuit will consist of a buck-boost converter, as system voltage is expected to both fall short of and exceed the nominal battery charge of 12 volts, and load matching is desired at a maximum range of operating conditions. A programmable microcontroller will be used to take thermocouple inputs and drive a pulse width modulation circuit that will serve as a switching regulator and control the performance of the DC-DC converter. The circuit components are currently in an early stage of prototyping and material lists are being compiled. Each bank of four thermoelectric modules in series will require a load matching circuit. There is some loss of precise load matching on a module by module basis in this arrangement, as each of the four modules will be at slightly different conditions. This is mitigated, however, by the facts that implementing a load matching circuit for each module would be almost four times as costly; instrumenting each module with thermocouples is impractical; and the conversion efficiency of a DC-DC converter to boost from a single module voltage to the desired 12 volt range would be much lower.

4.6 – Generator Section and Assembly

The piping work and custom generator section, as they are currently fabricated, required a large amount of welding, cutting, grinding, and other labor intensive activities to complete. The drawing package, including rough fabrication and assembly instructions, can be found in Appendix A. Overall, the craftsmanship and attention to detail are of a high quality, and the product is both sturdy and safe to handle. Interior weld interfaces, where accessible, have been ground relatively smooth and likewise the transition sections are of a tapered rather than abrupt nature in order to reduce the sources of head loss inherent to the structure of the system.

Despite efforts taken during fabrication to restrict the possibility of warping during welding, some warping of the base plate, to which the thermoelectric modules mate up, did occur. Judicious use of an acetylene torch and c-clamps was largely able to correct the problem, as it was primarily localized in one corner of the plate, however, it is evident that this fabrication challenge would be extremely difficult to mitigate with current techniques. If the design were to go into a mass production mode, the generator section would likely be assembled using extruded parts, thus greatly reducing the time and complication of excessive welding.

The final product presented, see figure 9 below, is the sum total result of a number of previous prototyping operations and as such itself presents opportunities for a number of further improvements. There are currently plans to significantly reduce the weight of the heat sink to enable the addition of a second generator subassembly without excessive impact to system weight. Additionally, while the assembly fits lengthwise in its install location, the clearances are less than optimal for installation troubleshooting. A set of relatively simple modifications will allow for a length reduction of an estimated 10 inches.

[pic]

Figure 9. Photo of assembled prototype.

4.7 – The User, Society, and the Environment

Once installed with the operation of controls verified, the thermoelectric generator should have very little direct impact on the user, as no user input and only very minimal maintenance will be required. For the typical user it will be yet another unseen automotive component. For society the impact will be improved fuel economy and perhaps a heightened awareness of vehicle efficiency issues. Impact to the environment will come by way of reduced emissions per unit of vehicle performance. With the exception of the thermoelectric modules, the materials of the system are not unique and should not result in any increased disposal problems. If mass production is achieved, it is possible that the semiconductor materials in the thermoelectric modules could require special disposal attention, but a recycling program seems possible as well.

5.0 – Product Evaluation

5.1 – Design Failure Modes and Effects Analysis (DFMEA)

DFMEA analysis was accomplished by evaluating the entire system, then concentrating on the subsystems, and finally the individual components. While numerous analyses were completed, only a limited amount of implementation occurred due to a truncated time schedule. Table 1 shows a summary of results; see Appendix D for details of DFMEA work.

|Component and Failure |Initial RPN |Corrective Action |Current RPN |

|Air ducting obstruction |90 |Put grill over scoops |30 |

|Hanger Bar the bar fails by |54 |Remove hanger bar by manf. heat |27 |

|breaking | |sink that bolts directly to | |

| | |exhaust pipe | |

|Heat sink has impact with |64 |Orient the heat sink on top of |32 |

|foreign object | |the assembly instead of the | |

| | |bottom | |

|Wire leads from thermoelectrics |128 |Trim and redirect wiring scheme |32 |

|snagged by foreign object | | | |

Table 1. Summary of major DFMEA design impacts.

The first corrective action taken was to orientate the heat sink and chips on top of the exhaust section instead of the bottom. The DFMEA analysis revealed that when the heat sink was hanging below the exhaust pipe an impact could cause the generator to fail or possibly fall off, resulting in a near total loss in terms of system function and replacement cost.

The second was to eliminate the hanger bars and angle irons that attached the heat sink to the exhaust pipe. DFMEA analysis revealed that the hanger bars and angle irons were an added liability that could become loose, fatigue, or otherwise fail and cause the generator to malfunction. Through manufacturing controls the heat sink was redesigned to attach directly to the generator section of the exhaust using fasteners.

The third was to route the leads from the chips on the inside of the heat sink instead of leaving them on the outside. The DFMEA analysis uncovered the potential for the wires to be entangled on roadway debris or tools during vehicle maintenance causing a major malfunction in the generator.

The last corrective action implemented was to put protective screens on the air intake scoops. The DFMEA analysis revealed that a potential existed for foreign material to become trapped in the air duct. With the implementation of screens this problem was largely eliminated.

5.2 – Testing Results

To date, experimental evaluation of the desired performance has not been achieved. This is largely due to the fact that the test stand arrangement using a hot plate, electric stovetop burner, or propane torch was unable to reach the necessary thermal conditions and to the fact that dynamometer testing of the Future Truck engine has not been possible. There are currently plans to conduct testing in concert with an FSAE or Clean Snowmobile dynamometer test run, but these results will not be available in time for documentation in this report.

However, it is possible to make use of some of the experimental data that was gathered on the test stand. The graph of figure 10 shows that the power output is indeed greatest at conditions approximating match load conditions (5.2 ohm). It can also be seen that while the measured data follows the same general trend as the theoretical performance, the theoretical curve fit is lower across the board for given thermal conditions. This may either be due to inaccurate temperature measurements during the experiment or it may be due to the fact that relationship describing output performance is only valid about the component design point (at a much higher temperature) and cannot be accurately extrapolated down to the testing temperatures.

[pic]

Figure 10. Experimental results of thermoelectric power output vs. temperature differential, with the black line representing a 5.2 ohm load resistor, the pink line representing a 11.8 ohm load resistor, the yellow line representing a 3.8 ohm load resistor, and the blue line representing the results yielded by modeling.

6.0 – Economic Analysis

6.1 – Expenditures to Date

Table 2 reflects a summary of expenditures to date. In general, the actual budget corresponds relatively closely to the originally proposed itemized budget. A number of items exceeded their estimated costs, however, a number of items were not actually implemented. Of primary importance is the fact that the load matching, DC-DC converters are not included above. Their recently projected cost of around $250 will bring the expenses more closely in line with the budgeted amount.

[pic]

Table 2. Condensed itemized expenditures.

6.2 – Labor Estimates

Table 3, below, is a slightly modified version of the labor estimate table found in the previous project report. While no concerted effort was made to log and track project time, resulting in very approximate times, the estimate seems fairly reasonable in retrospect. One difficulty of parsing project time into distinct activities is that it was a common occurrence for work to spill across category delineations. In the end, the most instructive aspect of the labor estimation is the fact that it is probably low and is billed at only twenty dollars an hour, yet it still amounts to approximately four and a half times as much as the cost of the hardware purchased. At a more realistic hourly cost, it is clear that for projects of this scale labor costs in the real world will far exceed materials costs.

[pic]

Table 3. Estimated total labor and tasking breakdown.

6.3 – Production Costs

Table 4 presents realized costs for the proposed expanded implementation scheme as compared to an estimation of costs in mass production at a major auto manufacturer. The listed costs for our project exclude costs of materials used in the prototyping process but not actually implemented in the final product and include the cost of an added generator set and load matching circuits.

[pic]

Table 4. System cost comparison: senior design vs. mass production.

Mass production costs are very rough estimates of economies of scale cost reductions. The highest degree of cost reductions are expected to be realized for the thermoelectric modules and the load matching circuits, both of which are currently specialty, small batch items that could benefit immensely by larger production runs. Other materials and hardware items will benefit from the preexisting supply relationships and in-house capabilities of an automobile manufacturer.

The power output listed for the senior design project is based on two banks of thermoelectrics and DC-DC converter efficiency of 88%. The power output of the mass production estimate is based on the implementation of the next generation of thermoelectric modules, which are projected to have thermal efficiencies in excess of four times as great as current modules. In addition, the conversion efficiency of the DC-DC converter is projected to be slightly higher.

6.4 – Case for Implementation

It is obvious that from the analysis presented in table 5 below that the product in its current state does not warrant implementation from an economic perspective, however, it should be remembered that the primary motivating forces behind the project were to provide an added feature of interest to the Future Truck vehicle for the purpose of the competition and to provide concept validation to encourage further development.

[pic]

Table 5. Projected return on investment for prototype vs. mass production.

Increased fuel prices and the introduction of the next generation of thermoelectric modules, the assumption of both of which is included in the mass production numbers, would be significant motivating factors toward further commercial pursuit of thermoelectric waste heat recovery in passenger vehicles. However, as foreshadowed by Paccar’s work, the most likely place for thermoelectric market penetration will be in long haul trucking where systems installation will not be as spatially limited as with passenger vehicles. Additionally, as fuel price plays a primary role in the comparison of initial capitalization cost versus lifetime fuel savings cost, it is likely that there would be a greater degree of interest in the full-scale application of thermoelectric waste heat recovery in Europe or other markets where fuel prices are, on average, much higher than in the United States.

7.0 – Conclusions and Recommendations

Given the investment made by Future Truck in this project thus far, the project hardware ought to be implemented on the vehicle as soon as the power train is reinstalled and operational. However, as there has been some concern with regard to the added head loss the system will entail, the system should not be installed in the vehicle if overall vehicle performance will be negatively impacted. Testing will be necessary to make this determination in a definitive manner. Flow modeling is obviously also an option to examine the head loss issue. Given the nature of the hardware, such modeling will have to take a computational fluid dynamics route. Initial efforts have been made toward implementing such a model, but this has not met with success due to the complete lack of experience with the software tools. If performance degradation due to head loss is not problematic, then the system ought to be implemented even if system output is minimal in order to showcase efforts to pursue unique automotive technologies. In a comparative sense, it appears very unlikely that the flow restrictions imposed by the thermoelectric generator section and the bypass valves will be greater than the effects of the extended overall length, dual two inch diameter piping, dual catalytic converters, and many rough transitions and welds found on the exhaust system of the previous Future Truck iteration.

It is clear that testing difficulties associated with creating representative thermal conditions outside vehicle installation testing as well as the lack of access to engine and vehicle testing in general were severely underestimated. The same could be said with regard to the lead time on various hardware acquisitions and with the resolution of the load matching circuit issue. Attempts at finding off-the-shelf products have proven entirely unsuccessful, with the majority of DC-DC converters specialized to very low power circuits and the rest being of a very specialized nature with unavailable tech support. However, recent coordination with two senior electrical engineering students has provided a strong lead for resolving the load matching issue. The work necessary is similar in nature to previous work done involving the ultra-capacitor bank, only on a smaller scale. Thus, with continued work, the critical topic of load matching should be resolved by mid-January.

In the design presentation a proposed second phase implementation budget was introduced. This called for the purchase of nine additional thermoelectric modules, materials for two new heat sinks, components for an additional two load match circuits, and some miscellaneous supplies relating to installation. The total estimated cost for this expansion is approximately $1,200, the substantial majority of which goes towards the thermoelectrics. As the next iteration heat sink design will weigh just over half as much as the current design, this expansion would provide a substantial improvement to system performance with minimal impact to system weight.

While funding and work should most definitely proceed at once on load match circuitry and a number of exhaust system components which are not directly related to the scope of this senior design project, from a pragmatic viewpoint it may be advisable to postpone final approval for the purchase of the additional thermoelectric modules until after positive in situ test results are achieved. Such testing should also give insight into the head loss issue that weighs upon the overall project impact. Currently, there seems little reason to think that the load match circuit issue will not be resolved successfully. The issue of when the authorization should be made for the purchase of additional thermoelectric modules then rests heavily upon lead time concerns. If CNC mill time is available, the heat sinks could easily be completed within a single week. Additional load circuits should likewise be quick to construct after the initial circuits are completed. The first purchase of thermoelectric modules took approximately three weeks to arrive after payment was made. Therefore, early February ought to be the deadline for additional purchase. Actual installation of the additional modules would be relatively expedient, as all preparatory work could be accomplished well in advance of the actual receipt of the modules.

With a projected output well in excess of 100 watts, an expanded thermoelectric generator would be able to power the onboard electronics and computers and offer the Future Truck a point of distinction.

Appendix A – Drawing Package

The drawing package was created using SolidWorks software. It includes as-built drawings for all of the major, in-house fabricated components used on the final product and also contains rough, step-by-step assembly and fabrication instructions for operations not directly discernable from the drawings themselves.

Dwg. No. Description

A1 Heat Sink Base

A2 Heat Sink Fin

A3 Heat Sink Assembly

B1 Box Duct

B2 Generator Subassembly

C1 Pipe Wye

C2 Generator Section

C3 Bypass Pipe

C4 Total Assembly

[pic]

[pic]

[pic]

[pic]

[pic]

[pic]

[pic]

[pic]

[pic]

[pic]

Appendix B – Thermoelectric Performance Model

This appendix presents a more polished, MathCAD version of the original Excel math model used to predict the power output of the thermoelectric modules for given thermal conditions, assuming a match load electrical condition.

[pic]

[pic]

[pic]

[pic]

Appendix C – Impact of Additional Weight Model

This appendix presents the first generation modeling used to estimate the relative importance of the weight the thermoelectric generator system will add to the vehicle. It can be seen that added system weight does impact engine load by an appreciable amount in comparison to system power output, but that this additional load should be well below the system output, resulting in a net gain. The accuracy of this model should not be overestimated, as the parameters are of a somewhat general nature.

Appendix D -- Design Failure Mode and Effects Analysis (DFMEA)

The purpose of the DFMEA is to define and guide a logical design process in order to identify, quantify, and reduce design risk. This appendix provides a traceable document for design and development in order to achieve continuous product improvement.

[pic]

[pic]

[pic]

[pic]

Appendix E – Testing

The operational procedures and the program used to collect the experimental data are outlined in the following appendix. The steps have been used with little change for the majority of the semester. The same general procedure would be used for a test on an actual exhaust section, where the hotplate is replaced as the heat source. It should be noted that the bench test described was used in parallel for both senior design and a senior lab project.

The Labview programs have constantly evolved through the year from a simple program that took temperature data to the program that is shown. The current program takes the data that is collected and shows it on the screen. In addition to that it calculates the power that is being produced by the chips, the theoretical power production, and then writes the data to a text file.

Also included is a graph of temperature data taken from the Future Truck exhaust during a city loop drive in August.

Operational Procedure:

1. Initially the computer needs to be turned on and Labview 4.0 software began running

2. The program needs to be opened and the sampling time inputted along with inputting the resistance that is being tested

3. The Peltier generator section must be places on top of the heat distribution plate that is set on top of the hot plate and clamped down

4. The hot plate needs to be turned on next along with the fan to blow ambient air through the heat sink

5. The Labview program has to be started up to begin recording data

6. The system has to stay like this until the hotplate reaches steady state

7. Once steady state is reached the hotplate is turned off and the Labview program is stopped

8. Once the heat distribution plate gets back to ambient temperature the resistor can be removed and the experiment can be reran with a different load

[pic]

Figure D1. Front panel of Labview program written for data collection.

[pic]

Figure D2. Wiring diagram of Labview program written for data collection.

[pic]

Figure D3. Bench test apparatus setup.

The above picture is of the setup constructed by PBJ Xaustors. The hotplate on the bottom can be seen and is the hot source for the chips. The fan on the back is the airflow source through the heat sink power off 12V DC. The ducting on the top is to keep the blown air flowing across the heat sink fins and allowing for the greatest cooling.

[pic]

Figure D4. Field Point hardware used for data acquisition.

Above is a picture of the Field Point modules that were used to acquire the data. The far left module is the control that exports the data via a CAT 5 crossover cable. The middle module has 8 thermocouple acquisition points, and the far right box has 8 voltage measurement points.

Figure D5. Thermal mapping of Future Truck exhaust during a city drive loop.

Appendix F – Résumés of Team Members

Updated résumés of team members documenting skills gained from the project. Robert Wiegers is entering graduate school for mechanical engineering at the University of Idaho. James Stewart is pursuing a career as a Naval Surface Warfare Officer. Peter Jorg will graduate in May, 2003 and plans to seek gainful employment.

James T. Stewart

1120 ½ S. Hayes, Moscow, ID 83843

email:stew5185@uidaho.edu

(208) 882-7453

PROFESSIONAL OBJECTIVE:

Research and Development Mechanical Engineer with Raytheon Service Company

EDUCATION:

· Bachelors of Science in Mechanical Engineering, University of Idaho-Moscow 2003

· Naval Science Minor, University of Idaho-Moscow 2003

· Navy Nuclear Power School Orlando, Florida 1996. Courses of studies included Mechanical,

Electrical, Metallurgical, Chemical, and Nuclear Engineering

· Navy Nuclear Machinists Mate “A” School Orlando, Florida 1995. Courses of studies included

Mechanical Engineering and practical applications

SUMMARY OF EXPERIENCE:

Highly skilled mechanical engineer and craftsman with more than three years of hands-on and

supervisory experience in the safe operation, maintenance, repair, and design of mechanical, hydraulic,

and propulsion systems within diverse engineering environments in the United States Navy. Vast

experience in quality control, ordering, and management of materials.

CAPABILITIES:

· Design, install, operate, maintain, repair, and overhaul industrial level steam, diesel, and nuclear

machinery including the following:

Pumps (centrifugal and positive displacement) Motors

Heat exchangers Turbines

Valves and piping Industrial Air Conditioning

Distilling plants Generators

Cooling systems

· Create schematics, technical drawings, and blueprints using CAD programs for fabrication, repair,

and maintenance of systems and components

· Supervise technicians in the implementation and completion of work packages

· Maintain complete and accurate records for work centers

· Ensure that safety regulations are carefully followed as regulated by the Nuclear Regulatory

Commission and the Occupational Safety and Health Administration

· Plan and coordinate production control schedules with emphasis on quality control

· Prepare and present technical reports to customers, resource specialists, planners, senior management,

and the public

ACHIEVEMENTS:

· Assisted Newport News Shipyard test engineers in performing hydrostatic tests on nuclear primary

and secondary systems prior to final acceptance of all work

· Managed and coordinated the efforts of shipyard, intermediate maintenance, and ship personnel for

the overhaul of the fuel oil transfer system, which was completed safely and ahead of schedule

· Supervised and trained 6 other Quality Control Inspectors

· Redesigned quality control packages for Technicians use saving approximately two hours per job

· Awarded numerous commendations for scholastic and military achievement

· Qualified as an Enlisted Submarine Warfare Specialist

· Completed Naval ROTC four-year training program

PROFESSIONAL EXPERIENCE:

Nuclear Maintenance Mechanic/ Emergency Nuclear Repair Welder, United States Navy,

July 1996- August 1999

Peter L. Jorg

Campus Address: 627 Elm St, Apt 207, Moscow, ID 83843 208-885-3967 E-mail:

Permanent Address: 1912 Lutes Rd NW, Poulsbo, WA 98370 360-779-5020 jorg5740@uidaho.edu

________________________________________________________________________________________________________________________________________________________________________________________________________________________

Education: University of Idaho Moscow, ID (Fall 2000 – Spring 2004)

B.S. Mechanical Engineering

Cumulative GPA: 4.0/4.0 Earned 100% of college expenses

Course Highlights

-Senior Design (Fall ’03) -Fluid Dynamics

-Advanced Engineering Graphics -Machine Component Design

-Heat Transfer -Macro/microeconomics

-Experimental Methods for Engineers -Thermal Systems Design (Fall ’03)

King’s West School Chico, WA (1996 – 2000)

Cumulative GPA: 4.0/4.0 Co-Valedictorian SAT: 1510 (800 Verbal, 710 Math)

Skills: -Solid Works/Edge, AutoCAD, MathCAD, MatLab, TK Solver, RISA, Word, Excel

-Basic Spanish, basic familiarity with Steel Code, some shop experience on manual and CNC milling machines

Projects: Senior Design. Two semester long project working in a team of three to implement thermoelectric power modules in the exhaust system of a hybrid vehicle to recover waste heat energy as electrical power. Experience gained in CAM product realization, working with a diverse group of vendors, understanding the challenges involved in bringing a complex project from concept design through fabrication and testing (ME 424 & 426, Summer & Fall ’03)

Dynamometer. Designed, built, and calibrated a dynamometer for the characterization of small DC motors in order to facilitate future ME 323 blimp projects. (Summer ’03)

Blimp Project. Teams of four designed, modeled, constructed and tested remote controlled, helium filled blimps powered by DC motors. In depth technical proposal, report and design documentation presented. (ME 323, Spring ’03)

Experience: Naval Undersea Warfare Center Keyport, WA (Summer, 2002)

o responsible for drawing-package analysis, organization, and emendation

o gained experience in CAD/CAM process in the realization of design modifications

o gained exposure to a wide range of activities on base and their interrelation

General Construction Company Poulsbo, WA (Summer, 2001)

o responsible for fabrication drawings, product research, material takeoff, design amendments to 100% planning drafts, and basic structural analysis

o gained experience in drafting, interpreting and communicating plans/drawings, problem solving, business parlance, using the Steel Code, and engineering analysis

Activities/ -Tau Beta Pi Engineering Honor Society -Phi Eta Sigma Honor Society

Awards: -University of Idaho Engineering Ambassadors -University of Idaho Scholar

-American Society of Mechanical Engineers -National Merit Scholar

Interests: -Music (6 years jazz band, 7 years youth symphony)

-Travel (American west, Western Europe, Canadian Yukon, Antarctic Peninsula)

-Writing and reading (Creative Writing class, fiction, history, political theory)

-Outdoors (hiking, rowing, archery, horses)

Appendix G – Project Timeline

This updated schedule displays steps leading toward project completion. The appendix compares the actual progress made during the semester and the initial schedule produced at the beginning of the semester. Each task has two bars associated with it. The upper bar represents the actual work completed and the lower bar represents the initial schedule produced at the beginning of the semester.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download