Proceedings of the Multi-Disciplinary Senior Design Conference Page 3 s5

Proceedings of the Multi-Disciplinary Senior Design Conference Page 3

Project Number: P10008

Proceedings of the Multi-Disciplinary Senior Design Conference Page 3

Arcworks fillvent line improvement

Andrew Lawlor – IE
Adam Janicki – ME / Chris Somers – ME
John Zeffer – ME

Abstract

The purpose of this project is to develop a pressure test fixture for the fillvent cap for Arcworks to improve their production line output by 10%. The total scope of the project also includes other areas of the production line in order to improve the output as well. The production line will be organized by 5S to improve overall production. The pressure test fixture will replace the current set up, which is applying air pressure to the cap and dipping it in water to see if air bubbles are present. The new fixture is automated and requires very little user input. The process is full inspection of every piece and therefore must be accurate. A vacuum fixture will be used to test the caps. Two parts will be run at simultaneously to improve the process time. The threshold value to determine a good part from a bad part is a certain pressure at a given time. The program will determine whether a part is good or bad and light a green LED for a good part, a red LED for a bad part, and a yellow LED while the test is running. This new fixture has decreased the cycle time by 8 seconds.

introduction (or background)

ArcWorks is a non-profit company that employs people with mental disabilities. The company has requested several projects from the Senior Design classes in the past to help improve the ergonomics and efficiency.

The design for this project was to implement an improved testing device as well as alter the assembly process to increase output.

Thermo-Fisher provided the current requirements for the testing of the ultrasonic weld. Based on these documents positive or negative air pressure was to be utilized.

The reason for the project is a result of currently utilizing a fairly crude and messy testing procedure. In addition, the current testing method is slowing down the assembly line.

Our project is based off of a general testing concept used by many manufacturing companies where sealing integrity is essential. It is a simplified and streamlined design to help decrease the complexity of the operation of the device.

process

Customer Needs

Previous to the start of the design process the customer was consulted. It was found that increased output on the line was a primary need. The company identified what was believed to be the bottleneck station on the assembly line.

The project team verified this by performing time studies on the operators. This was done through visiting ArcWorks multiple times and observing operators. However to ensure that an even baseline was kept for comparison only certain operators deemed by ArcWorks as being consistent were timed. This was to ensure that upon completion of the project, one could time the line again with the same operators to verify that improvements had been made. Once the bottleneck station was verified, the project was able to begin the design process. The first step in the design process is to gain insight from the customer on their direct and latent needs. These are explained in detail below.

ArcWorks identified many needs for the re-design of the water test station. The first primary need was safety. The machine needed to be safe for the operators to use. This was to be done through shielding any area on the machine with moving parts, if necessary, and keeping the operators hands free of any unshielded moving parts. Included in the safety, the machine needs to be ergonomically designed, keeping operators from developing any new medical conditions while operating the machine on a daily basis. The design must also be intuitive to the operator; specifically the number of steps taken to complete the station must be at a minimum and the results of the test should be clear and precise.

The machine must also perform at or above the level of the current system in place for finding leaks in the seal of the cap. Meaning that all parts found to be bad on the previous system must also be found to be bad on the new machine. The elimination of water in the process is a primary need identified by ArcWorks. Robustness of the design was also identified to be important. There are a variety of different part types that must be accommodated for in the design. Allowing the new machine to function with any part type is a necessary trait of the new design. Along with a robust design, the machine must be reliable. Maintenance should be minimal, and require few specialized tooling.

Design Process

Each team member to ensure that a variety of options would be examined generated design concepts. These concepts were then evaluated using Pugh’s Method of Concept Selection, and comparing each concept to the following attributes; ease of use, cost, dependability, test time, safety and output. Ultimately two designs scores were identical. Since each design was substantially different, the design process continued for each design as if it was the only one chosen. After further discussion with the customer it was decided that the vacuum fixture would be chosen with ease of use for the operator taking more of a precedence of the other attributes.

Figure 1. Proposed Positive Pressure Fixture

Machine Design

Shown below is a picture of the vacuum fixture. The cap sits upon a soft rubber to create a seal. The two buttons seen at the front of the fixture are for the operator to initiate the test. This will ensure that during the test the operators will not disturb cap thus invalidating the results. Shown behind the cap are the red and green light indicators. These indicators show the results of the test, red for a bad part and green for a good part. Below and behind the fixture is a differential pressure sensor to measure the speed of pressure drop and a vacuum to pressurize the fixture.

Figure 2. Finalized Vacuum Test Fixture

Figure 3. Components Affixed to Test Fixture

Operation of the machine is simple and intuitive for the operator to use. The operator will take a cap from their incoming lane. The operator will then place the cap onto the test fixture. Next the operator will press the start button. This will allow the machine to run unaffected by the operator. The vacuum will then run and determine if there is a good seal on the cap. If there is not a red light will indicate to the operator that the part is bad, if the part is good a green light will appear. The vacuum will then release its pressure and allow the operator to remove the cap and place in its respective bin. This process can be ran simultaneously both of the stations on the machine.

System Architecture

To allow the system to function autonomously from a personal computer interface, programming of the Programmable Logic Controller (PLC) must be performed. For this to happen a systems level understanding of the function of the PLC must be acquired.

The PLC will control all aspects of the machine and how it functions with the operator and how the machine determines a good part from a bad part.

Once a part is placed on the machine and the buttons are pressed the PLC will follow the same architecture for every part. The vacuum will be turned on manually and run continuously while testing is being completed. Once one of the pushbuttons is pressed, a normally closed valve will be energized and allow the vacuum to pull negative pressure from the cap chamber this process is indicated by a solid yellow light. The negative pressure in the cap is continuously measure by a corresponding pressure sensor and when the high threshold is met the valve closes and the yellow light will begin to blink. The test will run for a set time and the PLC will read the value of the pressure inside the cap. If the pressure is greater than the low threshold the cap will pass and a green light will turn on. If the cap fails it will be indicated by a red light. The system will wait until the buttons are pressed for the next part and begin an identical sequence. There also is a three position lever that determines the thresholds and test times for each part. There are three combinations of parts that are tested and each position is for each combination: Up = “Little-Little”, Neutral = “Big-Big”, Down = “Big-Little”.

Preliminary Testing

Once the machine was built to a usable stage testing of both good and bad parts was performed. This was done using a differential pressure sensor, the stand with rubber attached and the vacuum. The system was interfaced with a PC and data collection software, LabView. This software collected the pressure reading from the differential pressure sensor multiple times per second. Once the test was complete this data was then exported to excel.

The methodology used for the testing was to compare the time and rate of decay for good parts and compare them to the rate of decay and time for known bad parts. These parts were acquired from the customer after going through the current process.

Once a cap is placed on the rubber, the LabView software was manually ran. Immediately following the vacuum was manually turned on for approximately 3 seconds. Once this step was complete a valve was manually closed to seal off the cap leaving it pressurized. Then after 45 seconds LabView would be complete and visually show a graph of the decay and then the data was exported into excel. This was completed for approximately 50 bad parts. Once all the data was in excel a comparison between good and bad parts was observed.

Final Testing with Final Product

The next step in the testing was running the fully functioning machine in parallel with the current water test. Each part type was tested to ensure that the machine would work on all part types. We were provided with 200 samples of each combination of parts. Half of each of the parts were ran through the water test first and then the vacuum test, while the other half of the parts were ran through the vacuum test first and then the water test. This methodology was taken to see if one test had an effect on the other. Whether the part was good or bad for each test was collected.

Results and discussion

Water Test 1st
Part Combo / Water / Vacuum / Number of Parts Failed by Water but passed by Vacuum
Good / Bad / Good / Bad
Little-Little / 91 / 9 / 80 / 20 / 0
Big-Little / 62 / 0 / 58 / 17 / 0
Big-Big / 94 / 0 / 89 / 15 / 0
Vacuum Test 1st
Part Combo / Water / Vacuum / Number of Parts Failed by Water but passed by Vacuum
Good / Bad / Good / Bad
Little-Little / 97 / 12 / 73 / 36 / 1
Big-Little / 62 / 0 / 60 / 4 / 0
Big-Big / 90 / 5 / 73 / 22 / 1

The results of the testing showed that the vacuum test was rejecting more parts than the water test. Only twice was a part that was bad with the water test pass the vacuum test. There was roughly 100 parts for the water test first with the three part combinations as well as the vacuum first with the part combinations. With the water first there were no parts that passed the vacuum and failed the water. With the vacuum first there were two parts that passed the vacuum and failed the water. The vacuum test rejects more parts than the water because it eliminates the user variability looking for bubbles and the test is run for a specific amount of time, whereas with the water the specified amount of time is not always followed when the part is tested.

Machine Installation

The old assembly machine set up can make approximately 450 subassemblies per day. The new assembly line can theoretically output 600 parts per day. This is a 33% improvement over a single test of the seal with a pitcher of water.

Floor Redesign

Another objective was to redesign the floor layout to maximize efficiency of the line. This plan was not implemented. Seen below is the current and proposed layout on the left and right respectively.

Figure 4. Floor layout of the fillvent assembly line cell at ArcWorks. Left, before the floor redesign. Right, proposed floor redesign. The arrows in each layout show the work flow through the assembly process.

Parts follow a U-shaped path through the assembly process, moving directly from the end of one table to the beginning of the next. Also, the line now goes from right to left, as opposed to the opposite flow that was run in the previous assembly line. This allows the operators to maximize the efficiency of their brains spatial recognition area. Grasping parts on the right the operator is forced to use the right eye as the primary eye. This then translates to using the left side of your brain where special recognition is found. Once several iterations are performed of the operator grasping the parts in a right to left fashion the placement of the part to the next station will become natural. This will allow the operator to remove wasted time from the process by not needing to look where they will place the part. The increased efficiency of the operator may not be able to be measured immediately, however the team is confident that long term the results will be positive if implemented.