Improvements made since the break and what's to come
April 13, 2024 - By Hedda Grelz, Andres Calderon, Oselumenosen Ekhuemelo, and Adolfo Diaz
__________________________________________________________________________________
Work period of March 23 - April 13
The team’s original plan was to start this work period by integrating the pulley system into the vent so the SMA wire could be housed fully inside the vent case. After this, the team was planning on starting the validation. However, when the pulley system was implemented, the vent did not successfully open and close even though the team tried many different variations of assembly and adding materials to the fabricated pulley system to increase the slip of the wire. The team suspects that the main issue with the pulley system is that the scale was too small to implement moving pulleys, and the team instead had to implement small pins that the wire was supposed to rotate around. The very small diameter of these pins likely prevents the SMA from contracting properly when heated since the wire does not slip very well. The team put in significant effort to fix this but after much trial and error and to prevent significant disruptions to their overall Capstone schedule, the team decided to proceed to the validation process without the pulley system. While this tough decision meant the team re-scoped an important part of the project, they were still able to perform their planned validation process (described below) and prove the other parts of the vent design that were successful. Taking away the pulley system meant that the team needed to have the wire going straight down from the vent. This meant that the original battery testing case that was fabricated for validation could not be used since the attachment of the wire to the bottom of the case was not feasible due to the nature of the construction of the case. The team therefore built an alternative testing stand (shown later in this blog) to be able to attach the wire to a hook and proceed with the validation of the rest of the prototype that worked well.
After making this decision, Team Navy focused its efforts towards accomplishing the “Device Validation Completed” milestone from March 23rd to April 23rd for the SMA-Actuated Safety Vent for Lithium Ion Batteries. The team completed validation for SMA recoverable strain, maximum vent opening, closed vent position difference (resealability), activation time, active opening trigger point, vent height and diameter, and the overall weight of the vent components. The team did not manage to verify the passive opening temperature as the temperature sensor they have is too slow to capture the heat from the heat gun before the wire contracts. They also tried to find ways to measure the temperature of the wire but the options available were out of the team’s budget. The information below goes into great detail regarding the device validation and some challenges the team encountered.
Beyond completing the key aspects of the device validation, the team successfully concluded their last official Capstone presentation in front of their Capstone instructor Dr. Agrawal, and several other senior design teams. Team Navy’s sponsor NOV has also confirmed their intention to financially support the team. Dr. Song received $5,000 to fund this project and other senior design projects. This means that Team Navy will be refunded the amount spent on this project.
Succes of the Project
To assess the success of the team’s device, we first must review the goals that were set at the beginning of the year according to the literature and industry review of already existing solutions for LIB safety vents and requests from our stakeholders. Table 1 shows the target values for the team’s success criteria that were based on current vent industry standards, on-set time for thermal runaway, thermal runaway ignition temperature, and temperature rate of change at early stages of thermal runaway, and guidance from Dr. Song that will use the prototype for educational purposes. Along with the quantified values in Table 1 are also the validation measurement instruments and the current status of each measurement.
Table 1: Validation criteria used to measure the success of the project.
Some of the physical aspects of the validation for the safety vent are simple, yet very important. We managed to stay under the specified 500 grams. The assembled vent weighs 272 grams and 466 grams when adding the control system components (SSR, Arduino, and Breadboard). Both these values are under the target weight of 500 grams. Another important physical parameter was the vent dimensions (height and diameter). The target diameter of 50 [mm] was not met due to machining limitations, which is one of the larger short-comings of the team’s project. The initial assembly method for the vent part was threads between the vent case and the vent lid. However, the small size scale of the vent prevented the UH machinist from fabricating the parts with that method. The team also did not have the funding to look for other machinists who could include the threads and keep the vent the correct diameter. To fix this, the thickness of the vent walls had to be increased so the machinist could use set screws instead. This increased the diameter from 5 [cm] to 6.1 [cm], exceeding the team’s target value based on industry standards by 1.1 [cm]. The height target value was met with the vent measuring exactly 2 [cm], which is 0.5 [cm] less than the maximum and exactly at the team’s target. This dimension carries priority over the diameter since anything larger would cause problems with integration into the battery pack. The batteries used in common LIB packs are 6.5 [cm] tall and the battery pack is only 9 [cm]. This only leaves 2.5 [cm] of space left for the vent.
Apart from the vent diameter being too large, and the pulley system not being successfully integrated, the only other problem the team had during validation was not being able to measure the passive opening trigger point. The team had planned to record the temperature of the wire with an infrared laser temperature sensor or record the temperature of the air around the wire with the temperature sensor used for the control system. When attempting to get a reading during passive actuation, the team found that the diameter of the wire was too small to emit enough light for the sensor to read. The temperature used for the control system was a bit too slow to read the temperature increase caused by the heat gun used to simulate thermal runaway to get an accurate reading. Although the team was not able to measure it, manufacturer information specifies that the SMA wire’s transition temperature window is 68-72°C, based on the composition of the national (nickel and titanium). Fatigue validation is yet to be done, but the team has solidified a plan to perform this within the next week.
The remainder of the validation has been completed and overall been quite successful. The first part of the validation, the SMA recoverable strain, was validated last semester to set the team up for a successful spring semester. The goal was to verify that the wire had a minimum of 4%, but preferably 5% recoverable strain which means that the SMA can fully recover the strain and go back to its original shape upon heating. This was done to verify that the wire could be strained enough to get a maximum vent opening of 3 [mm] or more while keeping the wire short enough to integrate in the pulley system. The verification of SMA recoverable strain was done by hanging weights on the wire and actuating it to see the limit load it could lift. Both 4% and 5% strains could successfully lift 1284 grams with no plastic deformation. Figure 1 shows the setup used for this validation. Current from a variable power supply was used to actuate the wire and lift the load. After actuation, the load was taken off to prevent further stretching. The team could then verify that SMA returned to its original length for both values of initial strains.
Figure 1: Experimental setup to measure the maximum load and recoverable strain of SMA wire.
All other measurements were performed in the fully installed vent in the testing set-up. As mentioned earlier, the original battery case that was going to be used for validation was not able to be implemented without the pulley system. The team built an alternative testing setup that is shown in Figure 2. The active opening is shown to the left and the passive opening is shown to the right.
Figure 2:Validation setup for active opening (left) and passive opening (right)
The first thing the team measured in the assembled vent was the maximum vent opening. This was measured using a digital caliper that was placed on the vent lid and then pushed down as the vent was opening to its max position. 10 readings were taken each for the active and passive opening. The minimum requirement was 3 [mm] but the team goal was to get at least a 3.4 [mm] opening. The average opening for the passive actuation was about 3.3 [mm] and 3.5 [mm] for the active opening. Both cleared the minimum requirement, but only the active opening cleared the set goal of 3.4 [mm].
Vent resealability was also an important validation measurement, This is the vent’s ability to fully return to its initial position when closed after an actuation cycle. This was measured using a proximity probe connected to a multimeter. An initial voltage reading was taken when the vent was fully closed, the vent would then be activated and another reading would be taken when it closed. This was performed 10 times for each actuation. The team then converted the voltage values to [mm] using the calibration method taught in experimental methods to obtain a calibration equation. The difference between the initial and final state was calculated to see if the vent was fully closed or not. The team's goal was for 0 [mm] difference. Figure 3 shows the experimental setup used for this process. The result from averaging the 10 runs for each actuation method was a difference of 0.48 [mm] for passive opening and 0.44 [mm] for active opening. The values are not 0 [mm] but are pretty close. The vent lid has a thickness of about 1.2 [mm] so there is still no gap for air to enter the vent even though it did not fully return to its original position. The problem was likely due to the springs being slightly too weak, or because of the machining of the vent lid that makes it difficult for the vent to not get a little bit stuck when trying to close. The sealing of the vent could be fixed by adding an O-ring.
Figure 3: Experimental set up to measure voltage difference for closing states using proximity probe.
The team also measured the vent’s actuation time. For passive opening, the timer was started when the heat gun was applied to the wire and stop when the vent reached its fully opened position. The active opening time was measured from the moment the code was activated to heat the wire (SSR starts blinking), to when the vent was fully opened. Figure 4 shows the graph for both passive and active actuation times. The goal was for the vent to open under 10 seconds. The team managed to average a passive actuation of 2.10 seconds and an average active actuation of 1.38 seconds.
Figure 4: Plot graph of actuation time for passive and active openings
Lastly, the team measured the active opening trigger point. This was done by using the setup shown to the left in Figure 5. The thermocouple is initially at a room temperature of 22.5°C. It is then dipped into a cup with hot water at a temperature of 77°C. When dipped in the water, the thermocouple senses a temperature rate of change larger than 25°C/min. A PWM current is then sent through the SMA wire which heats up the wire and opens the vent. The code was set to print out the current temperature rate of change when the PWM code was activated. The results from three active openings are shown in the right part of Figure 5. This validates that the vent opens actively when the temperature rate of change exceeds 25°C/min.
Figure 5: Final active opening set up and PWM start condition code print outs.
Overall the team considers the project to be mostly a success, with a few short-comings. The premise of the project was to develop an SMA-Actuated Safety Vent for Lithium Ion Batteries. The main novelty of the project was the active opening that can actuate the vent at early stages of thermal runaway and thus enable earlier mitigation of pressure and heat. The team managed to make the mechanism work as intended under the specified actuation time. Most of the goals set at the beginning of the semester were reached during the validation process. There were only a couple of disruptions with the project goals. One of the main set-backs was not being able to implement the pulley system. The team should have started the investigation of this earlier since the small scale makes it difficult to implement successfully. The team believes that if this had been investigated earlier, or if the team had a budget to get custom-made moving pulleys for the pulley system, it would have been able to work successfully. The other main shortcoming was the compromise on vent diameter, which also could have been resolved if machining options had been investigated earlier or the team had a larger budget to pay for a machinist who could fabricate the threads in the original design. The active actuation time works as intended and is low energy which was one of our goals as well. Although there could be improvements to the project, the team is proud of the project and believes that with some more refinements to the design, this idea could be very useful in industry vent designs.
Improvements to the Project and Remaining Validation
With the remaining time available, the team will prioritize evaluating the fatigue life of the vent prototype through 5,000 cycles of operation. The team will implement a loop in the code for the active opening to iterate the actuation process 5,000 times. The team predicts that the SMA wire or the wire clamp will break first which would lead to the circuit of the wire heating being broken. If this happens, the Arduino will shut down the loop and output the number of iterations at which the prototype failed. This data will provide valuable insights into the durability and performance of the solution, allowing the team to pinpoint areas for improvement. The team will also investigate better ways to clamp the wire and see if there is anything that can be done to get the pulley system to work. Most likely, the time and money left will not be enough to make it work before the conclusion of the project but the team will put in effort to try a few more options.
Technical Poster Draft
Komentar