Stirling Heat Engine Generator - | ECE 445, Study Guides, Projects, Research of Electrical and Electronics Engineering

Download Stirling Heat Engine Generator - | ECE 445 and more Study Guides, Projects, Research Electrical and Electronics Engineering in PDF only on Docsity!

Stirling Heat Engine Generator

by Joseph Galambos Derek Wojtowicz Mark Sangern ECE 445 - Section H T.A. Bob Schoonover December 9th, 2008 Project # 2

ABSTRACT

This report details the design, construction, and testing of a stirling heat engine generator. The design of the generator is split up into four subsystems: the stirling heat engine itself, a permanent magnet (brushed) dc generator, a dc-dc converter to increase the voltage, and a dc-ac inverter designed to output a 60Hz, 120 Volt (RMS) quasi-sine wave signal, similar to what is supplied by a common household outlet. Although the power electronics circuitry operated as intended, an overestimation was made as to the amount of power that the stirling engine would be able to provide. As a result, the output voltage of the entire system was not able to rise consistently above 22 Volts (RMS) with a very small load attached, though it was able to operate at the correct frequency. The generator, however, is able to maintain approximately 22 Volts (RMS) for extended periods of time. A larger stirling engine should be able to supply enough power to get the desired output voltage, 120 Volts (RMS). With the output power and voltage from the stirling engine/generator combination being substantially less than anticipated, the power electronics circuitry had to be supplied with an external power source. The stirling engine/generator combination must output 12Vdc with a minimum output power of 0.75W to provide enough power for the power electronics to operate independent of an external power supply. ii

1. Introduction In order for the constantly increasing demand for electricity to be met, while preserving our natural resources as well as the earth as a whole, more efficient and alternative energy sources must be utilized. However, it is difficult to integrate alternative energy sources due to their characteristic of remoteness (i.e. the best wind resources are often located in remote areas, far from current load centers). For an alternative energy source to be desired it must be both economical and have the ability to retrofit with existing types of generation facilities; thus making it marketable. For this reason our group will design and build a stirling heat engine generator. The stirling engine has been around since 1816, however, not widely used. Engineers are beginning to incorporate the stirling engine into power applications, but they are still rare to come by. Our design will illustrate that these external combustion engines are efficient, safe, low cost and easy to operate, thus making them very plausible for the future of renewable energy generation. 1.1 Objectives The purpose of this project is to effectively (and efficiently) convert thermal energy into electrical energy via a stirling heat engine. The stirling engine has several advantages over conventional heat engines (such as the steam turbine). First, the stirling engine can achieve the highest efficiency of any heat engine and perform “theoretically” up to the full Carnot efficiency. Furthermore, the stirling engine is not limited to any one specific heat source. As a result, the stirling engine-generator setup can be retrofit with all existing heat engine-generator type power plants, utilized as a stand alone renewable energy system (running off of solar heat) or be used in combined heat and power applications which convert industrial waste heat into electrical energy. The stirling heat engine generator system will output a 120Vac, 60Hz signal that will be able to provide a continuous 10W of power. This configuration represents a per-phase equivalent of the three-phase generation system. Thus, it will be easy to scale up our results for single-phase residential (integrated or stand alone) alternative energy system, or three-phase utility generation. 1.2 Benefits  Can be implemented with any type of heat source  Constant output power (ideal for base load generation in renewable energy systems)  Low maintenance (engine mechanism is relatively simple)  Can be implemented in remote locations (ideal for remote military base generation)  Variety of sizes available from single-phase residential to three-phase utility  Higher efficiency than any other heat engine  Easy to operate (almost everyone can operate a controlled flame used for a heat source) 1.3 Features - 120Vac, 60Hz output - 10W of continuous output power - Heat sources can easily be interchanged

1.4 Block Diagram 1.5 Block Descriptions Heat Source: The heat source is the energy input for the entire system. It provides the temperature difference (between the working gas) needed to operate the stirling engine, and is not limited to any specific type. The fuel used to supply the heat source will be Ethanol. The output of the heat source (heat) is the input to the stirling engine. Stirling Engine: The stirling engine acts as the prime mover of the system, and will be manufactured by Böhm Stirling-Technik. It uses the temperature difference between its hot and cold end (of the working gas chamber) to establish a cycle of fixed mass gas consisting of four main processes: cooling, compression, heating and expansion. Thus, the stirling engine converts thermal energy (supplied by the heat source) into mechanical work. The shaft, or output, of the stirling engine will be fit to the rotor, or input, of the permanent magnet (brushed) dc generator. PM DC Generator: The permanent magnet (brushed) dc machine is the electrical generator for the system, and will be manufactured. It will be used to convert the mechanical work, done by the stirling engine, into electrical energy. The output of the dc machine will be a dc signal (since it is of “brushed” type) from the rotor windings/terminals, and provide the input to the dc-dc converter. DC-DC Converter: The dc-dc converter is the voltage regulator for the system, and will be specifically designed to meet the performance requirements of the system. It acts as a dc transformer, and will output a clean dc signal of 170Vdc (to match the peak phase voltage of the U.S. power grid). The dc-dc converter will also stabilize the output dc voltage at 170Vdc if the input voltage were to vary (due to variations in speed of the dc generator). The 170Vdc output of the dc-dc converter will serve as the input to the dc-ac inverter. DC-AC Inverter: The dc-ac inverter is the frequency regulator/control for the system, and will be specifically designed to meet the performance requirements of the system. It will convert the dc signal (from the dc-dc converter) to a 60Hz, 120VRMS signal in order to match the phase voltage and frequency of the U.S. power grid. The 60Hz, 120VRMS output signal from the dc-ac inverter will then be connected to a 10W load for measurements.

HEAT SOURCE

STIRLING

ENGINE

PM (BRUSHED)

DC GENERATOR

DC-DC

CONVERTER

DC-AC

INVERTER

Figure 2: Push-Pull Converter Schematic and PWM Pin Connections 2.4.1 DC-DC Converter Operation The operation of the push-pull converter is rather simple to understand and consists of three main steps. The first step is dc-ac conversion. This is done using power MOSFETs. The power MOSFETs act as switches, controlled by a pulse width modulator (PWM) chip. In order for an alternating current to be produced on the primary side of the transformer, the power MOSFETs must switch in such a way that a maximum of one FET is on during the switching period. Note that the transformer used in push-pull converters must have a center tap. The positive terminal of the input dc voltage is connected to the transformer's center tap. The FETs effectively open and close the connection between the outer transformer terminals and ground. As a result of the FETs switching, the flux through the primary coil is effectively alternating. The turns ratio of the transformer has a direct effect on the secondary (output) voltage induced. However, the duty cycle, D, of the active switches (i.e. the FETs) also has an impact on the secondary (output) voltage. Equation (1), shown below, illustrates this relationship: V (^) 2 V (^) 1

N 2

N 1

2D

Where subscripts 1 and 2 represent the primary and secondary side respectively. Using Equation (1), the turns ratio can easily be calculated for a duty cycle of 50% as shown below: V (^) 2 V (^) 1

The ac voltage induced on the transformer secondary must be rectified in order to convert it back to dc. This is accomplished using a simple four-diode rectifier.

2.4.2 DC-DC Converter Part Selection The physical requirements for each component must be determined to ensure safe and proper operation. It is clear from Figure 2 that the power MOSFETs must be able to block a forward voltage (i.e. forward blocking) of +12Vdc when “off”. Furthermore, the power MOSFETs must be able to conduct a forward current (i.e. forward conducting) in order to complete a current path when “on”. In order to determine the maximum current each switch must be able to conduct, the current on the primary side of the transformer must be known. The output power (10W) is known at the end of the ac-dc conversion process. Assuming each power electronics circuit is at least 60% efficient at a 10W load (for a safety margin), the power at the input of the power electronics circuitry (the dc-dc converter) can easily be calculated. Since the two power electronics circuits are in series their overall efficiency is simply the product of each individual efficiency as shown in Equation (2): = dc − dc ∗ ac − dc =0.60∗0.60=0.36 (^) (2) Finally, using the basic definition of efficiency (the ratio of output power to input power), the power at the input of the dc-dc converter can be calculated: 0.36=

10W

P INPUT

→ P^ INPUT =^

10W

=27.8W

Since the voltage at the input of the dc-dc converter is known to be 12V, the input current of the dc-dc converter can be determined using Ohm's law (result: 2.3A). Therefore, each switch must be able to conduct a current of 2.3A (or more). As described above, the power MOSFETs must be forward conducting, forward blocking for 2.3A and 12V respectively. The STB300NH02L N-Channel MOSFET meets these requirements, and will be the switching device used in this application. This MOSFET has a very low drain-to-source resistance, less than 2.2mΩ (when “on”), and a very low zero gate drain current, less than 10μA (when “off”) [3]. A similar analysis can be performed for the rectifying diodes. Diodes are naturally forward conducting, reverse blocking. Each diode must be able to conduct a current of 0.16A or more and be able to block a reverse voltage of 170V. The MUR1540 ultrafast-recovery diode is rated for 15A conducting current, and can block up to 400V [4]. Placing a capacitor across the rectifying circuit will reduce the output voltage ripple. It is obvious from the placement of the capacitor that it must be rated for at least 170V. The size of the capacitor needed is not so obvious. First an equation relating the capacitor to the desired voltage ripple must be derived. This can be done using linear approximation methods for current across a capacitor as shown below: I (^) C = C dv dt

≈ C

 v  t

can be estimated as the perimeter of the core (44.3mm). The length of wire for the primary can now easily be calculated (note that there are two primary coils due to the center tap): lenght of wire = 2 ∗ 20 turns ∗44. mm turns =1772mm=5.81ft To account for some error, 8ft of AWG-20 wire will be needed to wind the primary. Similarly, the length of wire needed for the secondary winding can be determined: length of wire = 285 turns ∗44. mm turns =12625.5mm=41.43ft Again, accounting for some error, 50ft of AWG-28 wire will be needed for the secondary winding. The final piece of the dc-dc converter is the PWM. The Motorola SG3526 PWM chip will be used [6]. For the push-pull dc-dc converter, pins 3, 8, 9 10, 11, 13, 14, 15, 16, 17 and 18 will be used. The following paragraph explains each of the pin functions and their connections. Pin 3 (Compensation) controls the duty cycle of the output waveform, setting this at a voltage level above +3.6V allows for a duty cycle of 50%. Note that the dead time (amount of time both MOSFETs are “off”) also effects the duty cycle. Furthermore, as described above in Equation (1), the duty cycle effects the output voltage. As a result, a variable resistor is connected to pin 11, allowing for manual voltage regulation during operation. Pin 8 is an active low logic input that shuts down the PWM. This pin will be set high (+5V) in order for the PWM to operate. Pin 9 and pin 10 control the frequency of operation. From the PWM datasheet [6] it was determined that for 10kHz operation RT=10kΩ and CT=0.02μF, these are then connected to ground. A +12V supply is within operating range of this PWM, therefore, pin 17 will be connected to the 12V supply of the dc generator. Pin 18 is a reference output voltage (determined by the supply voltage of pin 17) of +5V [6]. The input voltage to pin 14 comes from the 12V supply of the dc generator. Pin 16, and its complement pin 13, drive the power MOSFETs. The output reference voltage of +5V from pin 18 will be used as the +5V input to pins 3 and 8. Pin 15 will be connected to ground [6]. 2.5 DC-AC Inverter A quasi-sine signal is being used as the output waveform for this application because it is an acceptable alternative for low-power usage, and greatly simplifies the circuit’s design [6]. The inverter consists of a SG3526N pulse width modulator chip connected to two IR2110 High/Low side MOSFET Drivers which will control the gates of four MOSFETs arranged in an H-Bridge configuration. 2.5.1 DC-AC Inverter Pulse Width Modulator Selection The SG3526N PWM chip sends the signals that control the gates of the MOSFETs via the IR2110 chips. It is designed to output a square waveform with dead time in between the high and low signals; this will cause MOSFETs M1 and M4 (as shown in Figure 3) to be active at once, creating a +170V drop across Vout, a break (where the output voltage will be 0V), then cause MOSFETs M2 and M3 to be active,

reversing the voltage bias to -170V. Thus, an output signal roughly equivalent to a 120Vrms, 60Hz waveform will be produced. The oscillation frequency of 60Hz was determined by setting CT = 1μF and looking at Figure 8 of the PWM datasheet [7] to find the appropriate value of RT (these are timing components used to select the operating frequency of the PWM). It was determined that RT = 15kΩ. “Dead time” is a period which results in no output from the PWM chip for a period of time after a “positive” or “negative” output signal. Therefore, RD needs to be set to an appropriate value (to be determined through the use of a variable resistor) in order to ensure the RMS output voltage is 120V with a 170VDC input. The dead time period is determined using Equation (5): 120V=170V 8.33ms− T (^) DEAD  8.33ms

Figure 3: DC-AC Inverter Schematic The resistance, RD, must be found by trial and error to achieve TDEAD = 2.45ms. This will cause all MOSFETs to be inactive and the output to have areas of 0V for 2.45ms. Thus, the output voltage waveform resembles a quasi-sine wave (shown in Figure 4). Ideally, the period of the waveform to be produced, as chosen through the above parameters, will be 16.7ms, resulting in an output frequency, 60Hz. Figure 4: Quasi-Sine Wave 1 +Error 2 - Error 3 Comp 4 Csoftstart 5 Reset 6 - CS 7 +CS 8 Shutdown 9 Rt 10 Ct 11 Rdeadtime 12 Sync 14 Vc 15 Ground 17 Vcc 18 Vref outA 13 outB 16 15 kΩ 1 μF Rv 12 VDC 2 Com 3 Vcc 5 Vs 6 Vb 9 Vdd 10 Hin 11 Sd 12 Lin 13 Vss HO 7 LO 1 SG 3526 N IR 2110 12 VDC MUR 130 2 Com 3 Vcc 5 Vs 6 Vb 9 Vdd 10 Hin 11 Sd 12 Lin 13 Vss HO 7 LO 1 IR 2110 12 VDC

1. 47 μF MUR 130 M 1 M 2 (^170) From DC VDC/ DC converter M 3 M 4

• Vout - IRF 740 B IRF 740 B IRF 740 B IRF 740 B

1. 47 μF 1 μF

3. Performance Requirements 1. Output a 120Vac (+/-5%) signal at a 60Hz operating frequency 2. Output a minimum power of 10W 3. Obtain a minimum overall output efficiency of 25% 4. Testing and Verification Before actually implementing our project as a whole, the individual components (corresponding to each block, as described above) must be tested separately. Thorough testing of individual components is critical to the completion of the project. Individual testing allows for easier adjustments (by fine tuning individual components that are not performing as desired), and is essential for the debugging process. 4.1 Stirling Engine/DC Generator Test/Results The gear ratio for the stirling engine/generator coupling was determined to be 10.667 (with the radius of the stirling engine drive pulley at 3mm and the radius of the dc generator drive pulley at 32mm). This was done using a trial and error approach. The belt used to couple the two drives was a small rubber band (to further reduce the load on the stirling engine). Note the spacing between the two drive pulleys was such that the rubber band (i.e. belt) was tight enough to eliminate slipping. The final spacing was 8in (from the center of the stirling engine belt drive to the center of the dc generator belt drive). Table 1, shown below, illustrates the output voltage and power of the dc generator (while loaded with a 40kΩ load) comparing single-cylinder, two-cylinder and three-cylinder stirling engines. These results can be used to approximate the number of cylinders required to meet the desired 12Vdc output of the dc generator. Table 1: DC Generator Output Voltage for Multiple-Cylinder Stirling Engines VDC,OUT (V) Output Power (mW) Cylinders <1.80V - 1 3.54V 0.313 2 6.54V 1.426 3 Using the results from Table 1, a plot can be made, and is shown below in Figure 5, comparing dc generator output voltage to the number of cylinders. A trendline, extrapolating the results, has been added (shown in bold) to provide a rough approximation for the number of cylinders required for 12V operation. Note that since only three data points were used to obtain the plot, the extrapolation results may be inaccurate. To obtain precise results for the number of cylinders needed for 12V operation, further testing must be performed on stirling engines with four, five and six cylinders.

Figure 5: DC Generator Output Voltage vs. Number of Cylinders The result is a stirling engine consisting of 4 cylinders; however, the resulting output power is only 3.6mW. Therefore, a similar analysis must be done relating the number of cylinders to the total output power of the entire system. Of course, all of these results are for stirling engines of very small size. In order to obtain more realistic results, the size (i.e. volume of the cylinders) should also be taken into account. Figure 6, shown on the subsequent page, relates the output power of the stirling engine to the number of cylinders. The plot indicates that a stirling engine with 10 cylinders would only output 8mW of power. This indicates increasing the number of cylinders does not significantly increase the output power of the entire system. As described above, the volume of the cylinders strongly influences the output power. This is typical for all combustion engines. Note that the load used for testing the stirling engine/generator combination was the same for both tests (relating output voltage to the number of cylinders and output power to the number of cylinders). The output power of the entire stirling engine generator system is substantially greater than that of the independent stirling engine/generator subsystem. This is mainly due to the power supply used to power the PWM chips and gate drivers of the power electronics circuitry. The power supply activates the dc-ac inverter, whose input is supplied by the charged rectifying capacitor of the dc-dc converter, resulting in an output voltage waveform and thus output power (due to the attached load). As a result, the maximum output power of the entire system is greater than the maximum output power (or maximum load) of the independent stirling engine/generator subsystem. Output Voltage vs. Cylinders 0 2 4 6 8 10 12 14 0 1 2 3 4 5 Number of Cylinders Output Voltage (V)

The final test will be an efficiency test of the converter. This will determine how efficiency the circuit is performing. Calculation of efficiency is shown in Equation (8), where the input and output power of the converter are measured using a multimeter. =

POUTPUT

P INPUT

x100 (8) 4.2.1 DC-DC Converter Test Results The first test done on the entire dc-dc converter circuit was the Line Regulation test. As described above in the design of the dc-dc converter, a variable resistor is used to control the duty cycle. This allows for excellent line regulation (as well as excellent load regulation). For comparison, the Line Regulation test was performed on the dc-dc converter with no duty cycle control (the duty cycle was fixed at 50%, nominal for the 12V to 170V design), and with duty cycle control. Since the circuit is designed such that the PWM chip is powered from the dc generator, the input voltage would never be less than 12V (otherwise the PWM chip would not be provided with the proper voltage to function), thus, Line Regulation testing was performed for various voltages ranging from 12V to 15V. The results are shown below in Table 2. Table 2: DC-DC Converter Line Regulation Test Results No Duty Cycle Control Duty Cycle Control VIN (V) VOUT (V) VIN (V) VOUT (V) 12 178.6 12 170. 13 192.5 13 170. 14 205.9 14 171. 15 219.3 15 169. Using the results from Table 2 with Equation (6) the Line Regulation for both converters (one with duty cycle control and one without duty cycle control) can be compared: Line RegulationWITH CONTROL =

x100 (^) = 1.29% Line RegulationNO CONTROL =

x100 (^) = 23.94% It is clear from Table 2 that as the input voltage increases the output voltage of the converter with no duty cycle control will continue to increase. As a result, the Line Regulation would also continue to increase. However, as the input voltage increases the output voltage of the converter with duty cycle control remains constant. Thus, the Line Regulation would remain constant . The results indicate that the converter benefits substantially with the implementation of duty cycle control. The same can be seen from the results of Load Regulation testing.

Load Regulation testing was performed on both converters (one with no duty cycle control and one with duty cycle control), using an input voltage of 12V over various loads ranging from 1W to 12W. The results are shown on the subsequent page in Table 3. Table 3: DC-DC Converter Load Regulation Test Results No Duty Cycle Control Duty Cycle Control POUT (W) VOUT (V) POUT (W) VOUT (V) 1.60 195.61 0.80 171. 3.50 185.70 2.80 171. 4.40 183.95 3.60 169. 7.40 178.71 6.65 171. 11.30 170.38 11.20 170. 12.00 169.37 11.90 169. Using the results from Table 3 with Equation (7) the Load Regulation for both converters (one with duty cycle control and one without duty cycle control) can be compared: Load RegulationWITH CONTROL =

x100 (^) = 1.38% Load RegulationNO CONTROL =

x100 (^) = 15.43% It is clear from the results that the converter with duty cycle control maintains better load regulation than that of the converter with no duty cycle control. The final test performed on the dc-dc converter measured efficiency. As done with line and load regulation testing , the efficiency of both converters will be compared. Table 4, tabulates the results. Table 4: DC-DC Converter Efficiency Test Results No Duty Cycle Control Duty Cycle Control PIN (W) POUT (W) η (%) PIN (W) POUT (W) η (%) 2.53 1.60 63.12 1.68 0.80 47. 4.56 3.50 76.69 3.84 2.80 72. 5.52 4.40 79.78 4.68 3.60 76. 8.41 7.40 88.02 7.93 6.65 83. 13.55 11.30 83.40 13.81 11.20 81. 14.39 12.00 83.40 14.52 11.90 81. Using the results from Table 4, the efficiencies of both converters can be plotted versus the output power and is shown below in Figure 7.

This corresponds to a period of 16.55ms. Since the duty cycle, D, is 0.3544 and there are two instances of dead time per period, we can calculate an individual dead time using Equation (9): T (^) dead = T ∗0.5− D  (^) (9) Which gives Tdead = 2.41ms, very close to the intended 2.45ms. The above results for the PWM were achieved with resistances RT = 11.5kΩ and RD = 71Ω, along with capacitance CT = 1μF. Once the PWM is attached to the two IR2110 MOSFET gate drivers, and the MOSFETs configured in the H-bridge, the dc-ac inverter can be tested as a complete unit. Figure 8 shows the output of the inverter circuit with a 50VDC input (since none of the readily available power supplies can supply 170VDC). The result is a quasi-sine wave output at 61.18Hz with an RMS voltage of 42V. Figure 8: Inverter Output Waveform Testing the inverter with various loads and a 50VDC source yielded the following results, shown in Table 5. Ploss is the power required to run the power electronics (from the 12V source), and is fairly consistent for different loadings:

Table 5: Inverter Load Testing Results Rload (Ω) Pin (W) Vout,rms (V) Pout (W) Ploss (W) Efficiency (%) 350 5.034 42.11 5.018 0.494 90. 500 3.548 42.07 3.541 0.494 87. 630 2.804 42.07 2.799 0.494 84. 1400 1.282 42.09 1.283 0.493 72. 4300 0.413 42.15 0.423 0.498 46. 7500 0.233 42.17 0.24 0.505 32. 30000 0.061 42.13 0.061 0.498 10. The values of efficiency versus output power are plotted in Figure 9. It is apparent that as the output power increases, the efficiency increases. The devices used in this circuit could likely handle a load of around 1kW; therefore, if the trend in Figure 9 were to continue the converter would have an efficiency greater than 95%. Figure 9: Efficiency vs Output Power of DC-AC Inverter

Efficiency vs Output Power

0 1 2 3 4 5 6 Pout (W) Efficiency (%)