Project Progress Blog
October 11, 2024
This week the group began CAD modeling the column rotation limiting device and the main assembly for the rack and pinion. The group also began drawing the electrical schematic for the input encoding and motor control. We looked into the design of each specific component and discussed how that component would be manufactured. After this discussion we made certain small changes to part designs in order to make sure they are easily manufacturable.
Surveying the No. 34 car
October 25, 2024
The group began with finalizing the design of critical parts and begining technical analysis. The group completed technical analysis on gear stresses, bending stress, Finite Element Analysis, and electrical analysis. The group also began working on the project website. In our meeting with our advisor, Dr. Aziz, we discussed gears and the different ways to analyze gears.
November 8, 2024
The group completed analysis on the gears used in our project and verified that they would not fail in our application. The group also analyzed the keyway used to transmit the torque from the pinion shaft to the pinion gear. The group began testing a few different 3D-printed gears for the column rotation limiting device. The gears tested included PLA spur gears, PLA helical gears, and PLA herringbone gears. The group selected the herringbone gears to move forward with since they have the advantage of having more contact area which makes them stronger. Helical gears have the same benefit, however they push each other apart naturally and make an axial force. The axial force would need to be considered when choosing bearings and axial load-rated bearings tend to cost more. Conversely, the herringbone shape gives the benefits of more contact area while also canceling out axial forces. After selecting the herringbone gears, we printed them out of PETG to test, and then we printed them again out of PETG and made them 100% solid. This is a good finding, since 3D printing gears is more cost effective than off-the-shelf gears. Originally we planned to use commercially available metal gears, but they were going to cost a total of $60.
Electrical testing occurred relating to the speed control of the motor. MOSFET turn on transient was analyzed on an oscilloscope, using a gate driver under a variety of conditions. Initially this testing was conducted with a 60 W resistive load at 13.8 V, and then 24 V. Data was collected about the response with different duty cycles, and at different frequencies. The testing also included the effect of adding additional capacitance to the gate driver to assist with charge pumping. The ADC on the ATmega328p and the wires connecting the motor to the battery were also analyzed as part of the technical analysis.
Electrical Testing
November 22, 2024
Our group has began the manufacturing for our alpha prototype. We began with sourcing material. Some materials were ordered and others were found as scrap material at our university. A lot of manufacturing was started. We began with the steering column rotation limiting device and associated parts. We also manufactured the shaft coupler and parts to make the case. Many different manufacturing processes needed to be done for these parts. We utilized CNC plasma cutting, CNC milling, manual milling, manual turning, and TIG welding for this portion of our manufacturing. Many hours were spent in the SAE Baja garage using the manual lathe, manual mill, welder, and CNC plasma cutter. In addition, many hours were spent in the Stevens Machine Shop using the CNC mill and the manual lathe. Our group encountered a few challenges such as the machine shop only being available for use from 7am-3pm. For our CNC parts we utilized Fusion 360's CAM (Computer Aided Machining) feature to create toolpaths to tell the machines how to make our parts. Before manufacturing the case and case parts, we needed to design fixture plates. The purpose of the fixture plate is to hold our part in the vise during different setups. Two fixture plates were designed; one to hold the case for different operations, and one to hold the case cover for different operations. These parts were machined using a Fanuc Robodrill CNC which is new at Stevens. We had to learn how to operate this new machine as well as attend a training last Wednesday hosted by a representative of Fanuc. The new Fanuc machine will be good for the manufacturing of our Aluminum case for a few reasons. The first is accuracy, the machine being more accurate than the other machines means that we will be able to create precise interference fit bores for our bearings using interpolation with an endmill rather than having to use a boring bar. This saves time as well as removes the human error of using a boring bar. The next reason is that this machine is more powerful and has more features. The machine has 5x more horsepower, 5x more spindle speed, faster rapid travel speeds, faster cutting speeds, an automatic tool changer, and powerful flood coolant. All of this will allow us to reduce our cycle time, reduce human error, and have better surface finish. For example, one of the roughing operations for the case (the case has about 8 different operations) would have taken 1.5 hours on the old Southwestern Industries machine. On the Fanuc the roughing operation for the case was less than 20 minutes. This machine has countless other benefits compared to the old CNC milling machines. These benefits will be very important for our project especially considering the tight timeline for phase 3 of the project.
Further electrical testing was conducted to begin controlling the motor with multiple MOSFETs. A second gate driver was used to facilitate the simultaneous high-side and low-side operation. Different MOSFETs were used to study the affect of the drain current and on resistance verses the gate charge, which negatively affects turn on time. Further test with different model gate drivers, and paralleled devices to improve function. Development of the logic circuiting for the motor controller was also progressed, such as the anti shoot-through logic and the thermal protection circuitry.
Manufacturing the steering column assembly
Testing the motor on the No. 34 car
Manufacturing the shaft coupler
Manufacturing the rack case
Testing the MOSFETs and Gate Drivers
December 6, 2024
The finishing operations were completed for the case. The Southwestern Industries machine was used to bore the side of the case for the linear bearings. We used a boring head with long boring bar to ensure both bores could be done in the same setup. This is critical since it ensures both bores will be perfectly concentric. This is required since any misalignment will cause friction between the rack gear and the linear bearings. After the case was completed, the group moved to manufacturing the fixture plate for the case and the case cover. The same process and machine was used as the case in the Fanuc milling machine, in order to ensure accuracy and to ensure the parts would be made efficiently. The next part that was made was the pinion shaft. The pinion shaft was turned on the lathe from 1" 4140 stock. Then it was put into a Bridgeport milling machine to mill the keyway. An 1/8" precision ground endmill was used to ensure the proper size for the keyway. The final part was the rack gear. The gear was cut down on the horizontal bandsaw then it was faced to size in the lathe. Then holes were drilled and tapped into each end to attach to the heim joints. After these parts were manufactured, the prototype was assembled and we verified everything fit and moved freely without extra friction.
While work was still being completed on the motor control circuit, the microcontroller circuit was finalized and the PCB layout was designed. Since the board design was complete for the microcontroller, it was sent out for production. Once the team received the board back, the components were soldered and the microcontroller was tested to ensure it could be programmed on the board. For the motor control, the logic circuit to implement dead time was finalized and it was verified to be functional using the oscilloscope. The whole circuit was then prototyped on breadboard to ensure it could turn the motor. Tests to run the motor were successful, with the circuit able to run the motor in both directions and stop the motor near instantly. Since the motor controller was not yet fully integrated with the microcontroller, a small PCB consisting of several LEDs was designed to show a proof of concept for taking the steering input. The column potentiometer was connected to this board, and the microcontroller took input from the potentiometer and turned on corresponding LEDs on the board based on its position. This allows for the column to be rotated, and the team can verify the input by viewing the LEDs that are illuminated.
Finishing the rack case
Testing the steering input
Dead time testing
January 31, 2025
The semester began with fixing a problem with the alpha prototype. The rack gear the team ordered was not properly aligning with the pinion gear. At first the team thought there was a manufacturing issue with the machining of the case. After verifying the CAD model and the drawings from the gear supplier, it was discovered this was not the case. After measuring the rack gear, it was determined that it was completely out of specifications from the drawing and product description provided by the supplier. When the team contacted the supplier about this issue, they determined that they had received defective stock from their manufacturer. They sent us a correct rack gear a few weeks later and the team machined it to make the necessary modifications for integration with the assembly. The new rack gear is noticeably different from the incorrect rack gear and it works properly.
The team finished board layout for the motor control PCB now that the circuit design was completed and tested. The board layout was designed with heat dissipation in mind, since the circuit components may become warm during operation. The layout also includes a header for the connection of a BLDC fan in case it is necessary to remove heat once the PCB is contained in an enclosure. The gerber and drill files were generated for the PCB and they were sent out to a manufacturer for production. In addition to the work on the motor PCB, the motor was also tested electrically with the updated rack gear. The rack assembly operated smoothly with the motor and the replacement rack gear.
New rack gear vs. old rack gear
February 14, 2025
The team designed a few parts to aid in the testing of the steering rack. The beta prototype focuses on interfacing the system with the Baja vehicle. In order to do testing of the steering system the team designed three brackets which were welded together in order to attach to the rack case and connect to the motor. This allows for moving the steering rack with the motor independent of the Baja vehicle. The team tested the steering rack by lifting objects of known weights to make sure that the output force by the rack gear was the same as the calculated force. This is to make sure that there is no extra friction from any potential issues during manufacturing or installation. In testing a significant amount of weight was lifted however we experienced failure of the testing apparatus. The nylon rope being used to lift the bucket experienced tensile failure. The team performed a quick tensile stress calculation to ensure that a piece of ER70S steel TIG filler rod would work without failure. Using the filler rod more weight was lifted however then the group experienced failure of the bucket. The team will manufacture a suitable testing apparatus and test more weight.
The motor control board was received from the PCB manufacturer. The team soldered all the components to the board. The assembly process also included screwing heat sinks to the MOSFETs, first utilizing thermal compound to ensure sufficient heat transfer will occur. Once the assembly was completed, the board was tested independently without any other project components, using external 24 V and 5 V power supplies. The testing confirmed proper operation of the motor board and no visible defects. The control inputs were manually adjusted to turn the motor in both directions.
Manufacturing the bracket for the rack case and motor
Testing the motor with the bucket
Manufacturing reinforced bucket
Motor control PCB received and tested
February 28, 2025
The team began manufacturing some parts for the beta prototype. Last fall the team selected grease to lubricate the rack and pinion gear. Rubber bellows will be used in order to keep the grease from escaping the rack case and to keep foreign contaminants out of case. The team designed an aluminum piece which will bolt to the linear bearings and has a round profile with a groove to hold onto the rubber bellows using a hose clamp or zip tie. The team made these parts using square aluminum stock and held them with a 4 jaw chuck in the lathe. After ensuring the stock was centered in the chuck by using a dial indicator, the center of the stock was drilled and bored to make clearance for the rack gear. Then, the outer radius was turned to match the profile of the linear bearing flange. Then the round OD for the bellow was turned and the groove was added.
The team prepared to integrate the column and rack systems together electrically. Multiconductor cable was selected to connect the potentiometers to the microcontroller board. Further progress on the programming of the microcontroller was made, which will be tested with the prototype once the motor and feedback mechanisms are connected.
Wiring all the components together
March 14, 2025
The team used 1" 0.125" steel square tubing and 1/4" grade 30 chain to make a structure to support and hold the bucket from the steering rack. This method worked and the team was successfully able to lift greater than 184lbs, which was the calculated amount. The group designed a bracket to allow the system to mount to the Baja car. The group also manufactured a steel square tube which was then welded into the frame to interface the mounting bracket to the vehicle frame. Once the mounting brackets and tube were welded and installed, we were able to install the system into the Baja vehicle for testing. To test the full system in the Baja vehicle, the team made a steering wheel to attach to the steering column. A flat sheet of aluminum was CNC plasma cut to create the steering wheel and then the grips were 3D printed.
The team soldered cable to the potentiometers and connected it to the circuit boards. The power steering system hardware is now completely integrated, so the software to program the microcontroller was loaded on the chip. As testing went on the program was further adjusted to improve performance.
March 28, 2025
The team decided to change out the regular spur gears on the steering column with herringbone gears. The helical shape offers smoother and quieter operation while minimizing vibrations. However, the downside to helical gears is that they generate axial force which would require thrust bearings and could increase friction. The herringbone gear profile combines both the left and right handed helical gears on the same gear which cancels out the axial thrust forces.
