Name: Sam. Lastname: Bednarski. Address: 119 Rolling Ridge Rd. City: West Milford. State: NJ. Zip: Country: USA.

Transcription

1 Name: Sam Lastname: Bednarski Address: 119 Rolling Ridge Rd City: West Milford State: NJ Zip: Country: USA Phone: School hulman.edu School: Rose Hulman Institute of Technology GPA: 3.88 OutOf: 4.00 GraduationDate: May 2019 Major: Mechanical Engineering Professor0: Michael Moorhead

2

3 Mike Stanton Finish Process Engineer Ryobi Die Casting (USA), INC. (317) /28/2017 North American Die Casting Association David Laine Memorial Scholarship 2017 Application Supervisor Recommendation Samuel E. Bednarski Dear North American Die Casting Association, Sam Bednarski held the position of Finish Process Engineering inter this past summer at Ryobi Die Casting in Shelbyville, Indiana. He was a valuable asset to our team this summer bringing in new perspective, energy, and enthusiasm for engineering in the die casting industry. Sam delivered valuable results in our cost reduction efforts with his study and implementation of a spindle vibration monitoring system. He contributed design improvements to our part labeling printer stands that will greatly improve operator comfort and process line layout flexibility. Finally, he designed tooling for our laser etching system to improve the repeatability of our machine. I look forward to welcoming Sam back as an intern next summer, and into the die casting field professionally when he finishes his studies. Sincerely, Mike Stanton

4 Intern Experience at Ryobi Die Casting (USA) Sam Bednarski Finish Process Engineering Intern 6/17 8/17

5 As an engineering undergraduate student, nothing compares to the experience of working in a real business environment. I am very grateful that I had the opportunity to work at Ryobi Die Casting (USA) during this past summer of This was my first real experience of seeing how my education is applied to work at a company. Not only was I pleased with my experience as a Finish Process Engineering Intern at RDC USA, but I was equally glad to be able to use what I ve learned during my schooling and make a positive impact towards the company. Though it took some time to understand everything that was going on in the plant, one of the things I learned quickly is that the number one obstruction to productivity is unplanned downtime. This is an especially prevalent issue when it comes to machine failures, whether it be due to components breaking, programming errors, or others. Over the summer, my main project was focused around reducing unplanned downtime due to machine failure and therefore reducing production costs. In particular, the goal of the project was to implement vibration sensors into the spindles of deburr machines to allow early detection of an oncoming breakdown. The deburring process of removing any excess flash and sharp edges is one of the most important among the different procedures which parts goes through during the finish operation. However, it is also one which produces much downtime due to breakdowns. The cause of these breakdowns is largely due to rotational imbalance of the deburr spindle in the motor housing. When a spindle is assembled, its center of mass must be within a tolerable radial distance from the rotational axis to assure that no excess stress will be put on the spindle system. Unfortunately, there is no sure way to check this during the assembly process, and destruction to the spindle occurs regularly. This leads to unscheduled downtime where employees must remove and replace the broken spindle and the spindle must be sent out to be repaired. The spindle repair as well as the overtime of employees are the two major costs of spindle failure. This project was structured as a cost reduction with a scope of reducing the costs of repairing all deburr spindles by 12.5%, which based on past reports would be about $27,524 annually. I divided this project into three main tasks; collect a large amount of vibration data from a selection of cells, determine appropriate warning and damage output alarms, and install sensors for all deburr spindle housings. Though my time at RDC USA was short, I was able to accomplish the first two tasks completely, as well as set the path for the sensor installation. The bulk of time spent on this project went towards the first task of collecting data. Without understanding the normal operation conditions of the spindles, it is impossible to detect when abnormal conditions are occurring.

6 Over two months of data was collected from five different deburr spindles using a single vibration sensor. The sensor setup is shown in Figure 1. It continuously measures the vibration of the spindle, and reports averages over spans of approximately five minute intervals. Every day, the data was pulled from the sensor and converted to a visual representation to help identify trends in the vibration patterns. Figure 2 depicts one day s worth of data from a particular cell. It took me some time to be able to understand what trends were present in the data. Values near 0 mm/s were collected when the spindle was stationary for an extended period of time, and therefore does not provide any insight. Values between 0 and about 4 mm/s were collected when the spindle was stationary for a partial amount of time, and therefore also does not provide insight. The upper cluster is representative of the vibration velocity of the spindle during production. For the example in Figure 2, it appears to range from about 5 to 9 mm/s. As I continued to collect data, it became more difficult to link precise patterns and trends between different datasets. However, when data for each spindle is compiled into one visual, trends among individual spindles can be identified. Figure 3 shows all the data for a particular spindle, coded by the date of collection. For this spindle, the cluster of data appears to have a maximum around 11 mm/s with a few outliers. However, one will quickly notice the data points which rise significantly above this value, to almost 18 mm/s. This dataset surprised me when I compared it to the previously collected data, and sure enough the next day there was a brand new spindle in that particular deburr cell; the original spindle had broken down during the night. With this new information, as well as vibration collected from other spindles, I was able to make strong inferences on what reasonable values of warning and damage outputs would be. The final step in this project was to permanently install a sensor into every spindle housing. Though, since it took a relatively long time to find the correlations between the vibrations and operating conditions, this final task was not completed during my internship. However, I was able to make a substantial start to this task. The sensor in Figure 1 is attached to the spindle housing with a magnet. Though the magnet is strong, it was not uncommon to find the sensor in a different position than where it was placed the day before. On occasion it had fallen off completely. The sensors being permanently installed in the housings required a hole thicker than the housing wall thickness of the housing directly vertical to the rotational axis. The sensors needed to be installed on an angle, which required a drilling jig. The final part of the project which I completed was the design and fabrication of a jig which would allow for precise angled drilling. Just before I left RDC USA, a few holes were drilled and tapped using the jig, and sensors installed correctly. This process only needed to be replicated to the rest of the spindles at the time of my departure.

7 There were several small scale projects that I worked on during the data collection process of the vibration project. One was the design of a stand for label printers. The finish plant which I worked in at RDC USA was in the process of setting up production lines for new products. These lines included leak testers, and at these stations labels needed to be placed on every part. I was tasked with designing a stand for a label printer to be seated in so operators could quickly and conveniently take labels to place on to parts. I completed this project mainly through the use of the SolidWorks program, and converted my model to drawings so that it could be fabricated. The SolidWorks rendering of the printer stand is depicted in Figure 4. Similar stands already exist in the plant, however I decided to add a bit more functionality to my design. Some key features are the ability to change the height, depth, and tilt of the printer. This allows for the same stand design to be used for a variety of different stations which may be set up differently. Also, a dust cover was added to help protect the printer from the accumulation of dust. This will hopefully increase the lifespan of the label printer. Figure 5 is the fabricated printer stand installed at the leak tester station. The other major project which I worked on during my experience at RDC USA was the design of additional locators for part fixtures on a laser marking machine. Though many parts are still identified by label, some products are replacing labels with laser burns. Since this is a relatively new technique, there are many areas of improvement in its execution. One issue was the repeatability of the laser burn. Many parts would pass through the machine and receive a good burn. However, a significant amount would fail a readability test and have to be reworked or possibly even scrapped. The cause of this issue was believed to be the variability in how parts are loaded onto the fixtures. The way the fixtures were originally designed allowed the back of parts to rotate between several millimeters while loaded on the fixtures. Though this does not seem like much, the focal length of the laser is very precise and must remain accurately consistent for it to work properly. I was tasked to design additional locators to be attached to the preexisting fixtures. Like the printer stand project, I used SolidWorks to design the locators. This was especially useful since the parts being marked have complex geometries, even where the locators sit. Figure 6 shows the SolidWorks rendering of the original fixture with the two additional locators. The design of these locators were sent out to be fabricated, as well as the modifications to the fixture block. Towards the end of my internship, I was able to see the installation of one of the sets of fixture locators and see that it did reduce the number of bad burns by the laser marker for parts loaded onto that particular fixture.

8 After my insightful experience at RDC USA, I now have a new perspective when it comes to engineering. I dealt with many different challenges, several of which are not things that can be taught in the classroom. My internship experience has solidified my interest in the manufacturing environment, and especially the design aspect of engineering.

9 Figure 1. Data collection setup of the deburr spindle. The vibration sensor is attached to the top surface of the spindle housing with a strong magnet. A cable is connected to the sensor which dually serves to supply power and transmit data.

10 Figure 2. Data collected from a particular deburr spindle over one full day. The data is visualized as a scatter plot to aid in the detection of trends and patterns in the vibration velocity.

11 Figure 3. Data collected from a particular deburr spindle over many days. Each set of data spans one full day. This plot allows for the detection of trends away from those observed during normal operating conditions.

12 Figure 4. The final design of the label printer stand as rendered by SolidWorks.

13 Figure 5. The fabricated label printer stand installed at the leak tester station. The ability to change the height, depth, and tilt of the printer allows the stand to fit in many different setups. SolidWorks. Figure 6. Preexisting laser marker fixture with additional locators attached as rendered by