P14254: Underwater Thermoelectric Power Generation Team: Charles Alexander, Tom Christen, Kim Maier, Reggie Pierce, Matt Fister, Zach Mink Guide: Rick Lux
Agenda • DDR Objectives • Review Project Goals • Electrical Design • Mechanical Design • Manufacturing, Assembly, Test Plans • Bill of Materials • Risk Assessment • MSD II Project Schedule
DDR Objectives From this review, we hope to: • Ensure conduit/wire connections are sensible • Check that testing plans seem feasible
Project Goals • Demonstrate proof of concept of thermoelectric system • Use a temperature differential to charge a battery • Achieve maximum thermoelectric efficiency over a range of temperatures • Establish a UUV-based research partnership between Boeing and RIT
Electrical Design
Battery • Purchased HP 6-Cell Li-Ion Laptop Battery • 10.8V and 55 Whr • Li-ion expert expresses concern in charging method • Smart Battery Communication – Pinout – Road-Block
Battery Reasons for not being able to communicate with battery • Wrong pinout • Unable to communicate unless battery is active (i.e. charge and discharge) If battery communication fails… • Use a TI Fuel Gauge attached to battery • Charge/discharge battery conservatively
Battery Safety • Li-Ion are normally charged in a CC-CV modes • Increased risk involved with constant power – Overcharge – Battery Damage • Risk Mitigation – Confirm Simulink model – Software overvoltage protection in conjunction with HP protection circuit – Conservative Charge/Discharge controlled by software
DC-DC Converter From last time: Inverting Buck-Boost Minimum Output Capacitance: 101 mF
ZETA Converter Calculated Values: Cc = 350uF Cin = 245uF Cout = 280uF L = 200uH
SPICE Simulation With predicted component values: ~10mV output ripple voltage ● Adjusting components until maximum ripple obtained: 120uF
SPICE Simulation ● With 120uF capacitors, ripple voltage is ~28mV, independent of output voltage ● Selected 180uF Organic Polymer Capacitors with low ESR
Current Sensing Found a 24 bit ADC - Linear Technology Typical Resolution of Hall Effect Resolution - (Ultra low power consumption) Equivalent to: Following the same process, resulting measured power uncertainty is reduced to For the same resolution, i.e. , the resulting power dissipation is: For the same Hall Effect Sensor, rated power consumption/dissipation is ~12mW How does resolution affect data being read? 10 bit ADC, 5V reference: For a 1A signal from the sensor - equivalent to a current resolution or “uncertainty” of : In terms of power, this can be thought of as an uncertainty of:
Circuit Schematic 5V Regulator Off Page
Hall Effect Sensor
Controller & ADC
To Program the Controller... ● Sparkfun “Tiny AVR Programmer” ● Can be used for Prototyping ● Free!
DC-DC Transistors
Discharge Protection
Regulator (off schematic page)
MPPT and Controls Flowchart
Electronics Test Plans Connect system in normal operating configuration Connect thermoelectric DAQ system to thermoelectric output for input voltage recording Connect the DAQ to the I2C on the battery for output voltage recording Connect the DAQ to the output of the Hall effect ammeter Connect clamp meter around the positive lead of the thermoelectric Synchronize all data acquisition systems Check to ensure connectors are properly connect and working under safe conditions Start recording data, and start heater, start charging by turning on regulator Monitor data during charging periodically, and ensure still charging safely Charge complete Check state of charge, battery current, and battery voltage to ensure charging has stopped Turn off heater, stop recording data, turn off regulator Compile data, and calculate efficiency of power in and power out
Mechanical Design
CAD Modeling - Assembly
Exploded View Enclosure lid Conduit connection Clamping plate, bolts, and washers Heat spreader Top Insulation Heater Side Insulation Thermoelectrics Thermocouple connectors Heat Sink
Thermoelectrics Out of all modules surveyed, this one appears to be the best. We aren’t sure if the manufacturer specs will match actual performance Thermonamic TEHP1-1264-0.8
Thermoelectrics This module is available in the Sustainable Energy Lab Its properties and performance are known Thermonamic TEP1-1264-1.5
Thermoelectrics • Customer reqs were set assuming 4% efficiency. It is possible according to specs. 2x TEHP1-264-0.8 produce 18W with 450W in (4% efficiency) Drawbacks: • MPPT must draw minimum current at all times. • 2x TEP1-1246-1.5 would only produce 13W with 500W input.
Thermoelectrics 3x TEHP1-1264-0.8: • 500W in -> 15.6W out (3.12%) • 563W in -> 19.8W out (3.52%) 3x TEP1-1246-1.5 • 630W in -> 15W out (2.31%)
Primary Insulation • Ceramic Fiber Millboard Rated for λ=0.1 W/mK Compressive strength =12 Mpa (20% deformation)) • Need: Top – 2x (40 x 120mm) Long Side – 2x (58 x 120 mm) Short Side – 2x (58 x 53 mm) • Cost: $35 for 1 sheet • Supplier: Furnace Products & Services, Inc • Testing: – Applied a load of 175 psi, experienced 5.8% deformation (6.8 mm to 6.4 mm). – Thermal conductivity was tested to an average of λ=0.125 W/mK
Secondary Insulation • 3000F Ceramic Blanket • Rated for: λ=0.086 W/mK • Supplier: Cotronics Corp.
Clamping Overview TEMs, heating elements, and top primary insulation secured by bolting into baseplate
Clamping Analysis 110 psi of clamping pressure over TEMs: • 90 psi preload, 20 psi thermal load • 111 lbf preload per bolt, 25 lbf thermal load • E L = 0.13” • Relative Thermal Expansion: 0.003” • #8-32 NC 2A bolts yield an n f of 7.3
Clamping Hardware ● 8-32 Cap Screw ● Belleville Washer ● Flat Washer ● Nylon Shoulder Washer
Belleville Option • 3 Standard #8 Belleville Washers in parallel handle both load and deflection. • One heavy duty #8 Washer will handle load and deflection.
Bending Analysis ● 150 lbf loads applied to bolt holes ● Maximum deflection is 0.0002” - This occurs in the pressure plate
Baseplate Bending • Max deflection of 6.4 E- 5” in baseplate • This is less than the 0.001” specified by Custom Thermoelectric for acceptable mounting surfaces
Pressure Plate Bending 11 mm pressure plate of low carbon steel has a maximum deflection of 0.0002”
Pressure Plate Bending 6 mm plate of low carbon steel has a max deflection of 0.0013”
Clamping Assembly • Clean mounting surfaces using alcohol, lint free swab • Add TEMs to baseplate, using etched lines to locate • Add heat spreader and heater on top of TEMs with all sides flush • Add clamping insulation to heat spreader • Add Belleville and flat washers to bolts. Shoulder washers pressed into pressure plate. ● Insert bolts into pressure plate, use bolt holes to locate pressure plate ● Finger tighten bolts one or two threads to ensure proper engagement ● Use torque wrench to tighten bolts to 4 in-lb in increments of 1 in-lb following figure http://www.boltscience.com/pages/tsequence.htm
Enclosure Deltron 480-0080 • Comes with seal • Sealing surface is simple • Manufacturer provided CAD drawings Therefore: 225x148x104 mm • we should be able to ~$46 adapt it easily.
Heater McMaster-Carr - Model 3618K379 • 750W Cartridge Heater - $32.96 – 0.495” Diameter, 5” Length – Will use a variac and power analyzer to set the power to the desired 563W
Heat Spreader • Need an isothermal temperature across the surface of the TEM • ANSYS used to determine temperature gradient across top of TEM under 2 geometries for copper and aluminum: Analysis Walkthrough: https://edge.rit.edu/edge/P14254/public/Design/HeatSink/Heat%20Sinking%20Worksheet.pdf
Heat Spreader Results: • Copper – 40mm by 40mm = 0.3 deg C temperature variation – 40mm by 25mm = 0.8 deg C temperature variation • Aluminum – 40mm by 25mm = 2.1 deg C temperature variation • Acceptable temperature variation = 1.5-2 deg C – Will go forwards with 40mm by 25mm copper heat spreader due to small temperature variation and limited depth of enclosure (104mm)
Heat Sink Analysis • Needed to find convection coefficient in water – Modeled fins as flow between flat parallel plates – Heat sink will be oriented vertically to maximize convection – Convection increases as length of heat sink decreases
Heat Sink Birmingham Aluminium - Model 1850HS • 6063 T6 Aluminum (250mm wide x 150mm) • Cost: $140 (includes shipping from UK) • Resistance can be improved: – No modifications - 0.115 K/W – 25mm trimmed from each end - 0.079 K/W – 35mm trimmed from each end - 0.067 K/W Analysis Walkthrough: https://edge.rit.edu/edge/P14254/public/Design/HeatSink/Heat%20Sinking%20Worksheet.pdf
Heat Sink Before Trimming After Trimming
Cabling Conduit vs. Cable Glands • Only one penetration • Need one per cable (2) • Easier to change wires • Cheaper • Might spill water into other end.
Recommend
More recommend