SYSTEMS DESIGN REPORT ELECTROMAGNETIC PROJECTILE PROPULSION DEVICE REV. 3 COMPLETED BY: NOAH WALKER COLLABORATION WITH: SAMITH CHOWDHURY GREGORY PAYNE PHILIP BURRELL SUBMITTED TO: UNDER GUIDANCE OF: DR. JONATHAN BACKENS EENG 498: ELECTRICAL ENGINEERING CAPSTONE (WI) PHYSICS COMPUTER SCIENCE AND ENGINEERING (PCSE) DEPARTMENT CHRISTOPHER NEWPORT UNIVERSITY 4 DECEMBER 2021 TABLE OF ABBREVIATIONS DC Direct Current LED Light Emitting Diode LiPo Lithium Polymer PCSE Physics, Computer Science, and Engineering PPE Personal Protective Equipment EPPD Electromagnetic Projectile Propulsion Device Li-ion Lithium Ion CV Computer Vision PSU Power Supply Unit LIST OF FIGURES Figure 1 Table of Important Considerations Figure 2 System Overview Diagram Figure 3 Table of projectile Design Alternatives Figure 4 Table of Capacitor Design Alternatives Figure 5 Table of SBC Design Alternatives Figure 6 Table of Hardware Design Components Figure 7 Wiring Diagram for Charging, Discharging and Propulsion System Figure 8 Relay Circuit Schematic Figure 9 Voltmeter Circuit Schematic Figure 10 Voltmeter Code Snippet Figure 11 Relay Control Code Snippet Figure 12 Plate Shatter Experiment Figure 13 projectile Velocity MatLab Script Figure 14 Capacitor Energy Available MatLab Script 1 1 INTRODUCTION 1.1 Problem Statement and Context An electromagnetic projectile propulsion device (EPPD) is to be built for this capstone project. The device uses the Lorentz force to launch projectiles at high velocities. This project aims to solve the size and portability constraints of the current rail guns developed by the navy. The project will be a proof of concept rather than a final product due to the legal limitations of making a firearm. The main use case of this project would be in personal use in the military, or on mounted turrets on land or air vehicles. Railguns are currently still in the research and development phase with no massive leaps learned in them other than more power, although this project aims to provide a new avenue to begin looking down. This project is presenting the same concept in a smaller form as a potential point of interest. 1.2 Requirement Analysis The key requirements for this project include the EPPD being portable, detecting targets to fire at, and having the ability to fire a projectile with enough force to shatter a ceramic plate. For safety the final step to fire the projectile would be the user confirmation for the user to confirm they are aiming at what they want to shoot. Two kill switches would be used,one to kill the propulsion and one to kill the charging systems. A voltmeter would also be included to verify that the systems have been fully turned off and discharged to allow safe handling of the turret. 1.3 Important Considerations Multiple considerations were made relating to the magnitude of the project. The impacts considered include public health, safety, welfare, global, cultural, social, environmental, and economic factors; the impacts are summarized in Figure 1 below. The impact to public safety would result from a new device entering production that could be untraceable in crimes, additionally the same point could be brought forth that it could be used to prevent crimes and protect the country. Ethics of war and weapons will not be discussed in this project. The impact to safety would result from improper usage and handling of the device. The high voltages used in the EPPD could result in instant death, or serious damage. Additionally the EPPD could severely injure someone if they were hit with the projectile. The impacts globally would come from an increase in military funding pushing forward more jobs and potentially raising tensions. Additionally, they could come from an increase in military power/technology which would also increase tensions. 2 The social impact would be from a political point of view with both sides of the United States political system viewing projectile propulsion devices/guns rights and laws differently, this will not be discussed in this report.. The impact to the environment is from the zero emissions shooting of the projectile. Along with the potential to power the weapon through renewable energy, although it could also be powered via coal and other non environmentally friendly The economic impact of the project could be an increase in military spending for the research and development of portable EPPDs. There are no welfare and cultural impacts. Factor Applies (Y/N) Brief Discussion Public Health Yes If it were to enter production, it could be used to save lives Safety Yes Misfires or improper handling, risks can be mitigated with protection equipment and controlled environment and operation Welfare No Does not apply Global Yes If it were to enter production, it could be sold to other militaries Cultural No Culture does not apply Social Yes Potential concern about new weapons Environmental Yes No emissions Economic Yes Increase in military spending for research and development 3 Figure 1: Table of Important Considerations 2 SYSTEMS DESIGN 2.1 Systems Overview The EPPD is separated into four sections.All the sections are connected via the control system to allow them all to function together. The sections include the target acquisition system, propulsion system, charging system, discharging system, and the control system. The target acquisition system includes a single board computer with software running to identify targets and give instructions to the motors based on the location of the targets to move the turret into position using the X and Y direction DC motors. The propulsion system includes the two rails to house the projectile and provide the connections and stability to enact the Lorentz force on the projectile. Along with a firing pin and spring to enact the initial velocity onto the projectile to prevent arc welding. The rails are linked to a capacitor bank which is in the charging system. The charging system consists of a capacitor bank, charging circuit and a battery to power the other systems. The charging circuit consists of a 24VDC - 360VDC DC to DC boost converter which takes the 24V battery in and outputs 390V to power the capacitor bank and charge it. The discharge system consists of a relay and a resistor bank to bleed out the voltage currently contained in the bank in the event of not wanting to fire the EPPD. The control system consists of a voltage monitor, relays and an LCD display for the charging system and the controls for the motors and operation of the propulsion system. The relays are used to change between different modes of operation such as charging, discharging and firing. Figure 2 Below illustrates the overall design and functionality of each system. With yellow being the control system, red being the discharge system, green being the charging system, blue being the propulsion system, purple being the target acquisition system and the gradients between two controls representing things that are used in both or control both systems. 4 Figure 2: System Overview Diagram 2.2 Design Alternatives and Constraints 5 C-Shaped Washer Regular Washer Steel Ball Bearing Cost The cost of the washer would be a minor portion of our budget. Negligible difference of cost between alternatives. The cost of the washer would be a minor portion of our budget. Negligible difference of cost between alternatives. The cost of the bearing would be a minor portion of our budget. Negligible difference of cost between alternatives. Feasibility Would have a stronger contact with the rails due to several points forming a line. Would have good contact to the rails with the double the weight of a c-shaped washer. Would have a single point of contact to the rails. User Friendliness Least user friendly since alterations would be needed to cut the projectile into a C shape. No alterations would be needed although this is harder to load into the barrel since there is less human grip than bearing. Most user friendly since no alterations would be needed and easiest to load into the barrel. Reliability Zinc is a weaker metal, will withstand less than steel would. Zinc is a weaker metal, will withstand less than steel would. Stainless steel is a stronger material than Zinc, thus letting it withstand more. Manufacturab- ility No manufacturability concerns for off the shelf products. No manufacturability concerns for off the shelf products. No manufacturability concerns for off the shelf products. Ethical Consideration No ethical considerations with the type of projectile. No ethical considerations with the type of projectile. No ethical considerations with the type of projectile. Limiting External Factor(s) Weight, price, conductivity, and strength. Weight, price, conductivity, and strength. Weight, price, conductivity, and strength. Changes with budget alterations If budget allowed, custom manufactured C shaped washers would be purchased. If budget allowed, a stronger and more conductive material washer would be purchased. If budget allowed, a stronger and more conductive material bearing would be purchased. Figure 3: Table of Projectile Design Alternatives 6 Several projectile types were looked into as shown above (Table 2.1), to see which would provide the best firing capabilities, the c-shape projectile was the final. This is due to the c-shape projectile having good contact with the rails through the whole process while also being economically friendly to the project with 1 washer resulting in 2 c-shape projectiles. We decided against the washer because it is simply a waste compared to the c-shape projectile with it being 1 projectile, instead of 2 c-shape projectiles. The steel ball bearing was also not a feasible option after some research and finding that the ball would melt and lose contact with the rails through the firing process, not delivering all the power that was intended to the projectile. With a larger budget we could have custom milled out projectiles that would fit perfectly with our rails to provide aerodynamics along with the correct contact points. Jianghai Cornell Dubilier Electronics Rubycon Cost $34.50/Unit $121.08/Unit $6.41/Unit Specifications 450 VDC @ 5600μF 450 VDC @ 5600μF 450 VDC @ 270μF User Friendliness Screw in tops ensure proper wire connection Screw in tops ensure proper wire connection Normal soldering points. Reliability Reliability should be fine with voltage within specified range. Reliability should be fine with voltage within specified range. Reliability should be fine with voltage within specified range. Manufacturab- ility No manufacturability concerns for off the shelf products. No manufacturability concerns for off the shelf products. No manufacturability concerns for off the shelf products. Ethical Consideration No ethical considerations with the capacitor. No ethical considerations with the capacitor. No ethical considerations with the capacitor. Limiting External Factor(s) Price, specifications, and strength. Price, specifications, and strength. Price, specifications, and strength. Changes with budget alterations This is the best option given the budget. Overpriced unit for similar specifications to Jianghai Better price with worse specifications compared to Jianghai. Figure 4: Table of Capacitor Design Alternatives Several capacitors were looked into as shown above (Table 2.2), with the 450VDC @ 5600 uF at $34.50/unit being the final choice. This choice was made due to the $121.08/Unit set containing no difference that pertained to the project and was significantly more expensive than the chosen 7 capacitors which left less room for other purchases. The $6.41/Unit capacitors were ruled out due to the need to solder the capacitors, and with the voltage and current we are going to want, we preferred to have industrial grade connections to the capacitors rather than a simple soldered point. Money would not have made much of a difference in this situation if we had more, the only difference being that we could have chosen a different brand of capacitors that could be considered “better”. Nvidia Jetson Nano Raspberry Pi 4 Rock Pi n10 Cost $59-179/Unit $35-95/Unit $99-169/Unit Specifications Quad-Core 1.8GHz Processor. Memory up to 4 GB. Upgraded GPU. No Neural Processing Unit. Quad-Core 1.5GHz Processor. Memory up to 8 GB. No upgraded GPU. No Neural Processing Unit. Dual-Core 1.8 GHz Processor. Memory up to 8 GB. Upgraded GPU. No neural Processing Unit. User Friendliness Challenging interface and feasibility, Very basic and easy to use, past experience Challenging interface and feasibility Reliability Will supply sufficient graphic processing power Not enough graphic processing power Will supply sufficient graphic processing power Manufacturab- ility No manufacturability concerns for off the shelf products. No manufacturability concerns for off the shelf products. No manufacturability concerns for off the shelf products. Ethical Consideration No ethical considerations with the computer. No ethical considerations with the computer. No ethical considerations with the computer. Limiting External Factor(s) Memory Limited graphic processing power Price, strength Changes with budget alterations Purchased outside of allocated budget Purchased outside of allocated budget Purchased outside of allocated budget Table 5: Table of SBC Design Alternatives Several SBCs were researched when considering how to operate the targeting control system. Above is a table describing different specifications (Table 2.3). We decided on the 4GB Nvidia Jetson Nano due to its significantly upgraded Graphics Processing Unit (GPU) which is the main element used in object detection. The second design choice was the Rock Pi n10, which has similar specifications. Cost was the deciding factor, and decided the Neural Processing Unit 8 included in the board would not be necessary because the software to be implemented doesn’t require neural networks. Other options did not meet performance requirements with respect to object detection. More money in this situation would not have made much of a difference with the current board already being fairly top of the line for image processing. 2.3 Engineering Standards WORK IN PROGRESS 9 2.4 Hardware Design Hardware Usage Elegoo UNO R3 Control relay system to turn on and off individual systems, control firing servo motor Nvidia Jetson Nano Computes the CV code and relays information to directional servo motors 450VDC @ 5600 uF Capacitor Charges and discharges into rails Aluminum Rails x2 Used to hold projectile and create Lorentz Force C-Shape projectile Used as the shorting mechanism between the rails DC to DC Boost Convertor Takes 24V 5A input and outputs 390V 0.2A output to charge capacitors 5V Elegoo Relay Module Terminates connection between Li-ion Battery and DC to DC Boost Convertor Lexitron 4KHD Webcam Allows the CV script to see in front of it 24V Li-ion Battery Powers the system 400V 0.2A Relay x2 Turns on and off discharge and charge system 100W 2.2k ohm Resistor Bleeds out the voltage in the capacitor Servo Motors x2 Used to change the direction of the turret Servo Motor Used to pull back pin for firing PLA Filament Used to print Barrel, and Firing Pin Figure 6: Table of Hardware Design Components The table above lists all of the included components in the creation of the EPPD and what their uses are from a hardware perspective. 10 Figure 7: Wiring Diagram for Charging, Discharging and Propulsion System The figure above, Figure 7, shows the schematic for the charging, discharging, and propulsion system along with some of the control system as shown by Vref_relay and the charging and discharging switches. The charging system is made up of the 24V DC 5A battery which is connected to the DC to DC boost converter, and controlled by the vref_relay which controls and powers the relay to allow the connect the DC to DC boost converter and the 24V DC 5A battery. The DC to DC boost converter then leads into the charging switch which while in an on state with the discharge switch being off will charge the capacitor bank. When the charging switch is off and the discharge switch is on the capacitor bank will bleed out its charge through the 2.2kOhm 100W resistor. In the event that both switches are off the capacitor bank will turn the rails on and will wait for the projectile to short them and fire the projectile out of the barrel. 11 Figure 8: Relay Switch Circuit Schematic The figure above, Figure 8 shows the wiring of the Elegoo UNO R3 to the 30VDC 10A Relay, which consists of a wire from the GPIO 1 pin of the Elegoo UNO R3 to the IN port on the Relay to control the open and closing of the relay. A wire going from the 5V pin on the Elegoo Uno R3 to the VCC pin on the 30VDC 10A Relay to provide power to the system and allow the relay to function. A wire leading from the GND pin on the Elegoo UNO R3 to the GND pin on the Relay to ground the whole system and a battery to provide power to the Elegoo Uno R3 to let it function. 12 Figure 9: Voltmeter Circuit Schematic The above figure, Figure 9 shows the voltmeter circuit schematic which contains a set of resistors R1 and R2 which are connected to C1 which is our 450V 5600uF capacitor. The values of R1 and R2 will be tweaked to provide the best measurement while staying to a minimal current draw to prevent an overdraw from the DC to DC Boost Converter and to prevent a massive bleed of the capacitor. There is then a wire leading from the junction point of the two resistors to the A1 pin on the Elegoo Uno R3. This port can only take 5V max input, meaning that the values of the R1, R2 voltage divider will have to drop the voltage reading to about 1% of the current voltage in the capacitor. There is then the DC to DC Boost Converter which charges the capacitor, and a battery to power the Elegoo Uno R3 board. 13 2.5 Software Design The software used to complete the prototype was Arduino IDE and PyCharm. The Arduino IDE was used to develop the software for the Elegoo UNO R3 while PyCharm was used to develop the CV script that runs on the Nvidia Jetson Nano. val = analogRead(A0);//reads the analog input Vout = (val * 5.00) / 1024.00; // formula for calculating voltage out i.e. V+, here 5.00 Vin = Vout / (R2/(R1+R2)); // formula for calculating voltage in i.e. GND if (Vin<0.09)//condition { Vin=0.00;//statement to quash undesired reading ! } Serial.print("\t Voltage of the given source = "); Serial.print(Vin); delay(1000); Figure 10: Voltmeter Code Snippet The code shown above in Figure 10 is run on the Elegoo UNO R3 in conjunction with a breadboard containing two resistors of known value to read in the voltages, it contains a squash if statement for low undesirable voltages. That statement will be removed in the future when we measure higher voltages. 14 if(readline(Serial.read(), buf, 80) > 0) { if(atoi(buf) == 5){ relayState = HIGH; Serial.write("High"); } if(atoi(buf) == 0){ relayState = LOW; Serial.write("Low"); } } digitalWrite(5,relayState); Figure 11: Relay Control Code Snippet The code shown above in Figure 11 is run on the Elegoo UNO R3 in conjunction with the 4 Channel Elegoo Relay Module. It initializes the pins and reads in user inputted values which will be changed to read the measurement from the voltmeter. It then turns on the relay in a LOW state and turns it off in a HIGH state. 3 PROTOTYPICAL IMPLEMENTATION 3.1 Implementation Details The current implemented hardware is as follows. The Elegoo UNO R3 has been implemented and connected to the Elegoo Relay Module to terminate the connection between the power and the DC to DC Boost Convertor which has also been implemented. The PLA Filament has been implemented to create housing for our rails and firing pin. The C-shape projectile has also been implemented along with the 450 VDC @ 5600 uF capacitor which has been connected to the rails. The implemented software is currently the code snippet from Figure 8 and Figure 9 in conjunction with some additional code to properly measure values from the voltmeter and trigger relays. The tools used to implement the above mentioned were a 3D printer to use the PLA Filament and create the housings we needed for components. A screwdriver to tighten the connections on the relays to make sure the wires have a good connection, and to tweak the output value of the DC to DC boost converter. A multimeter was used to correctly measure the voltage coming out of the boost and verify that the Elegoo based voltmeter was functioning correctly. A 25V power supply was used to power the DC to DC boost converter in the absence of the 24v Li-ion batteries. Arduino IDE was used to compile and flash the code onto the Elegoo UNO R3 to read the voltmeter and control the relays. A saw and pliers were used to create the C-shape projectile by cutting the washer in half and then folding it to the desired size. 15 3.2 Testing Details We tested the energy needed to break a plate by dropping a 500 gram weight from designated heights using a meter stick to measure the heights accurately. The process is shown in Figure 12 below. Figure 12: Plate Shatter Experiment We then took these measurements that had been collected from this experiment and plugged them into the formula below. Using the known variables of 500 grams for the mass and 9.81m/s for the gravity. (1) 𝑃𝐸 = 𝑚𝑔ℎ The outcome of this plate drop experiment was a fairly accurate number on the force needed to break a plate when firing our projectile. It was found that the 500g weight had to be dropped from 32cm to give us enough energy to shatter the plate. Plugging this into Eq. 1 with the known values discussed above we find that the energy needed would be 1.6J to shatter the plate. This information was then plugged into the MatLab script shown in Figure 13 16 % How fast do we need to shoot a projectile? % Mass of Projectile (Washer cut to C-Shaped projectile) projectile_mass_g = 2; projectile_mass_kg = projectile_mass_g/1000; % Kinetic Energy Formula reworked for velocity % V = Sqrt(KE/(0.5*m) meters per second projectile_velocity_ms = sqrt(plate_shatter_energy/(0.5*projectile_mass_kg)) projectile_velocity_fts = projectile_velocity_ms*3.281 Figure 13: Projectile Velocity MatLab Script This code snippet above Figure 13 used the Eq. 2 below to calculate the velocity needed on the projectile to reach the desired energy of 1.6J. (2) 𝑉 = 𝐾𝐸 0.5*𝑚𝑎𝑠𝑠 With these calculations it was found that the velocity needed to reach the 1.6J energy threshold was 40m/s. After a conversation with the Dahlgren engineers that worked on the railgun produced by the navy we were told that most EPPDs of this size have 1% -2% efficiency, this meant instead of a capacitor capable of generating 1.6J we would need one capable of generating 160J. With this information we used Eq. 3 to determine the potential energy in a capacitor. (3) 𝑈 = 1 2 * 𝑐 * 𝑉 2 17 % How many capacitors of some given size will result in the velocity % Capacitor 390V 5600 uF capacitance = 0.0056; %5600 uF voltage = 390; %V % Capacitor potential energy % U = 1/2*c*v^2 capacitor_energy = 0.5*capacitance*voltage^2 % Capacitor Quantity needed capacitor_quantity_conservative = capacitance_tofire_conservative/capacitor_energy capacitor_quantity_liberal = capacitance_tofire_liberal/capacitor_energy Figure 14: Capacitor Energy Available MatLab Script We balanced efficiency and budget to find capacitors that could be used to fire the projectile at the speed we needed, and we came to the conclusion that a 450V 5600uF capacitor would be the best choice. We plugged it into Figure 14 and found that we needed 37% of a single capacitor to propel the projectile at the desired speed. We tested the implementation of the charging and propulsion system by using mock tinfoil rails, the 24VDC 5A to 390VDC 0.2A Boost Converter and a 5600uF 450V Capacitor. The system was wired up to a 25V 1A Power Supply Unit (PSU) and a Multimeter. The 25V 1A PSU was connected to the DC to DC Boost Converter. The output of the DC to DC Boost Converter was taken to the positive end of the capacitor and the negative end of the capacitor was attached to the negative rail. The negative rail was then connected to ground which was also connected to the ground on the DC to DC Boost Converter. After charging the capacitor to multiple values ranging from 100V to 250V a projectile was dropped with forward momentum onto the tin foil rails. The outcome of the testing was concurrent with what we expected, the tin foil could serve as a crude method to apply force but incredibly inefficient in doing so with it either vaporizing, or not being able to deliver a good contact between the rail and projectile. The circuit components used in the testing worked as expected the DC to DC Boost Converter only went up to 386V but that is well within the range we need so that value is not an issue. The capacitor performed incredibly well with it dumping all of the charge instantly when the projectile hit the rails. One slight issue 18 was the boost converter leaking some residual charge into the capacitor after it had been fired and turned off. This can be fixed with our final system with the relays not allowing any residual charge to flow into the capacitor after it has been shot. 4 CONCLUSION 4.1 Results and Outlook From the project requirements in our statement of work, we have achieved the following, ● Killswitch to disable any electricity and propulsion controls ● Killswitch to disable software and object detection processes ● Voltmeter in the capacitor bank for monitoring charge ● User confirmation as the final step in firing the device (Human-in-the-loop model) The remaining to be achieves are the following, ● Laser pointer fixated on the barrel to show aimed direction and current target ● Multi-colored LED light that will display the current state of the device (ex. “Charged”, “Scanning”, “Dormant”) ● Manual control option in the event that object recognition is not needed by the user ● Clear interface through VNC Viewer that allows for the user input as well as displays specifications of the device ● Signal tone activated by the final user confirmation to alter others around. Future work for the project will be finishing the propulsion system fully and beginning work on the remaining project requirements listed above along with finishing the discharge system. The outlook of the project is good, with all software components nearing completion and the charging system being over 50% done, all that is left is a propulsion system with target acquisition and some safety measures and data collection measures to be implemented. 4.2 Lessons Learned 3D Printing parts worked well for the most part with us being able to choose the dimensions that we wanted and exactly what we wanted to be printed. Although exact measurements are hard to achieve, so we needed to reprint a couple times to get the desired spacing that we went for. Next time we would space our tight areas of the print by 2-3mm on each measurement to guarantee it can fit into the slot without forcing it. Simulating parts was a vital component, with the capacitors being simulated and measured to ensure that we could get the charge and discharge in a reasonable time. Along with knowing the values for our capacitors to be able to give the energy needed to fire the projectile. One issue 19