Abstract
The bearing life test fixture is designed to collect data while a ball bearing is running through its theoretical life cycle. Data will be collected through the use of an accelerometer, which will continuously collect information while the bearing is under load. Loading conditions include heating the bearing to 230°F, applying an axial load of 360lbs, and rotating the bearing at 79.8RPM. The outer race of the bearing is fixed, while the inner race spins. The fixture is currently not operational. The fixture is currently not operational. Future work can be performed to correlate data taken from the fixture with the actual life cycle of the bearing.
Nomenclature
CAD(Computer Aided Design)-the use of computer technology for the process of design and design documentation
DAQ (Data Acquisition) - converts analog signal into digital data
Fuser Roll-a roll that essentially “fuses” the image to the paper so it cannot be altered. Found in the Xerox iGen line of printers
iGen-Xerox iGen Digital Press. Production color printer manufactured by the Xerox Corporation
LabVIEW-a graphical programming environment used by millions of engineers and scientists to develop sophisticated measurement, test, and control systems. Developed in 1986 by National Instruments Corporation
Solidworks-a 3D CAD Software package that includes simulation tools.SolidWorks is a DassaultSystémes S.A. brand.
Xerox Corporation–headquartered in Norwalk, CT, it is the world’s leading enterprise for business process and document management
introduction
Printers are complex pieces of machinery that include many parts that all move together to get the most precise image quality possible.The Xerox Corporation has been a leader in the printer and copier industry, paralleled by none in terms of quality and durability. The fuser roll in the iGen unit sees many different forces and loads as the printer runs its cycle. The heat rolls apply heat to the fuser, while the pressure roll applies the force needed to cure the image to the paper. All these loads are seen by the bearings that support the fuser roll, one on each side of the roll.
Figure 1: Schematic of Fuser Roll Assembly
BACKGROUND
Xerox Corporation approached our group to help them with a solution to a bearing problem they were experiencing in their iGen printers. As the fuser rolls come back for remanufacturing every 200,000 prints, Xerox Corporation had no way of telling if the bearings were able to be reused. As a result, many bearings that could have been reused were scrapped, leading to wasted money and resources. The team’s goal in this experiment was to design and run a fixture that can characterize the life of the bearing under the conditions seen in the printer. These conditions were given to us by the members at Xerox Corporation. The bearing experiences anexternal temperature of 230°F while operating, as well as an axial load of 360lbs. The outer race is fixed, while the inner race is spinning at 79.8 revolutions per minute.
Using these conditions, the team was told to design a fixture that could put the bearing under the given conditions as seen in the printer. Because a fixture like this was not in Xerox Corporation’s possession, the team used the help of Dr. Stephen Boedo, and his expertise in bearings. Dr. Boedo has done a lot of research in the field of bearings, and has published manypapers on this subject. With the support of the Xerox Corporation, alongside the help of Dr. Boedo, the group had to come up with a way to characterize the life of the bearing.
Design process
- Project Specifications
A list of specifications was created by the group and supported by the customer before concept generation started. This ensured that the group and the customer were on the same page as to what was being designed and built. These specifications included loading conditions, error of measurements, as well as measurement specifications.
Table 1: Engineering Specifications
*Unfortunately, cost of the torque meter was over-budget, and this metric was consequently eliminated from the final assembly.
- Concept Generation
The team started the design of the fixture by using a whiteboard to come up with ideas on how to achieve the specifications set forth by the Xerox Corporation. This included how to load the bearing, how to heat the bearing, how to spin the bearing, etc. After the ideas were written down, the team then came up with four different options that could be feasible. From there, a Pugh Selection Matrix was created and used to compare the different setups. In the end, the team chose to look at each option and take the best parts and compiled them into the final fixture setup.
The final design had the following characteristics:
- Fixed outer race
- Pneumatic loading
- DC motor
- Belt driven shaft
- Conduction to outer race
- Accelerometer for measuring vibration
- Thermistor for controlling heat
- Microcontroller
- LabView interface
With this information we were able to draw up a crude initial design with pencil and paper:
Figure 2: Concept drawing of the fixture
- CAD Design
Refined construction of the fixture design was completed using SolidWorks. This software package allowed the team to quickly and accurately model the whole fixture assembly. It also allowed for quick changing of dimensions as the assembly process took place.
Figure 3: Final proposed design of fixture
SolidWorks enabledthe team to see what the fixture would look like, and how it all fit together. Not shown above is the pneumatic setup, only pictured is the pneumatic cylinder. Also missing from this CAD assembly is the electronic components and other various pieces of hardware that will work together to make the fixture run.
The accelerometer can be seen attached to the top of the upper shoe, with the bearing sandwiched between the two shoes. The lower shoe is supported by the “pancake style” pneumatic cylinder, and is allowed motion up and down only by the guide rails on either side. Further down the shaft, the support bearings help to hold the shaft level, as well take the force of the moment created from the pulley at the end of the shaft. The motor is not connected directly to the shaft because the team wanted to isolate the vibrations caused by the motors as best as possible, so Dr. Boedo suggested that a belt be used to drive the shaft.
- Pneumatic Loading
The team decided to load the bearing via a pneumatic cylinder that pushes up on the bottom shoe that surrounds the bearing. Some formulas were used to calculate the size of the pneumatic cylinder, as well as what pressure was needed. The pressure coming out of the wall was 70psi, so the design intent was to stay below this to ensure the bearing would see a continuous load of 360lb.
Equation 1: Area of cylinder needed for 70psi
Equation 2: Diameter of cylinder needed for 70psi
Because the team wanted to stay under the 70psi, a 3.00-inch bore pneumatic cylinder was chosen. This required the need to calculate the new pressure needed for 360lb.
Equation 3: Area of 3.00in bore pneumatic cylinder
Equation 4: New pressure for 360lb with 3.00in bore
Because the pneumatic cylinder did not need a long stroke to load the bearing, the Parker “pancake style” pneumatic cylinder was chosen. It has a 3.00in bore with 1.00in stroke, and comes in just under 3 inches tall. The connector has a 5/8”-18 female connector, for which we bought a 5/8”-18 rod to connect to.
Figure 4: Parker Hannifin “pancake style” pneumatic cylinders
To compliment the pneumatic cylinder, a solenoid valve was also purchased. Because the pneumatic cylinder is double actuating (which means it uses air to extend as well as return), a 4-way/2-pos valve was needed. This will be controlled via LabView. Also included in the pneumatic setup is a semi-precision dial regulator, so that the pressure can be regulated to the 50.9psi needed to load at 360lb. Lastly, a filtration system was necessary to keep debris that might be present in the air supply from getting in the cylinder. The system is plumbed with thermoplastic tubing and push-to-connect fittings for easy installation.
- Driving the Shaft
A large number of bearings will need to be tested for the customer to get the data they require. The team wanted to make sure that the fixture would not fail due to fatigue. The main area of concern was moment created by the alternating load created by the pneumatic cylinder.
Equation 5: Largest moment generated by the shear stress.
Equation 6: Stress created by the largest moment.
The max stress for infinite life of 1045 Steel was calculated to be 5515.56 psi making for a factor of safety of 1.585 for infinite life.
Equation 7: Maximum stress for infinite life.
*A factor of 1 was used eventhough the shaft could be seeing temperatures of up to 230°F because at that temperature Kd would be greater than 1.
Equation 8: Factor of Safety
- Heating the Bearing
It was necessary to heat the bearing in the test fixture to simulate accurate operating conditions in the iGen machine. The team determined that the operating temperature of the bearings in the iGen machine is approximately 230 degrees Fahrenheit. Based on this specification, an appropriate heating mechanism was chosen.
The heater chosen for the project was a custom-build Mica Band Heater from Nordic Sensors Industrial Inc. It was configured in such a way that the heater sits between the outer race of the bearing and the shoes apply the load to the bearing. The heater was to be attached to the bearing shoes with J.B. weld, which was chosen due to its strong bonding capabilities, even under very high temperatures. Choosing a 120 Volt heater (line voltage) at the maximum available 60 watts. The heater interfaces with LabView to regulate temperature.
Figure 5: SolidWorks Simulation results – No barrier
A thermal simulation was performed in SolidWorks to determine if the heater was adequate to heat the bearing up to 230 degrees. The simulation determined that this heater would work, but it must be isolated in some way from the bearing shoes to avoid acting as heat sync, robbing the temperature form the bearing itself.
Figure 6: SolidWorks Simulation results – With barrier
A thermal resistant coating Zirconium dioxide was selected for the shoes to prevent heat transfer from the heater to the shoes. Unfortunately, time constraints prevented the team from having the shoes coated.
NOTE: In the final iteration of the project, it was decided to eliminate the heater. For more information, please see the section entitled “Results and Instructions.”
- LabView Interface
The entire assembly is controlled via a LabView interface providing ease of use and a platform that is industry standard. All signals are collected and sent via a National Instruments DAQ that handles both the analog inputs of two thermocouples and the digital input of an encoder. The device allowsLabView to collect all the pertinent information on the system's operating conditions and respond using the digital outs on the DAQ. In terms of signal processing the program will collect several time intervals and average them to create a Fast Fourier Transfer (FFT). This FFT graph will show the intensity of vibrations throughout the frequency domain. From this any peaks can be investigated and compared against the expect wear signals. In addition to data processing the LabView code enables automation of the system starting with the operation of the motor and pneumatic cylinder. This entire system will be handed over to Xerox Corporation along with documentation on how to operate it as well as notes on how to modify the code.
Figure 7: LabVIEW Interface
Results and discussion
-Heating the Bearing
The band heater used did not have a uniform thickness as expected. The thickness of the heater varied between 0.200” and 0.250. This prevented a tight fit between the bearing, saddle, and heater, which would have adverse affects on the vibration data.
Figure 7: Mica Band Heater
Figure 8: Comparison of Bearing Saddles
-Motor Coupling
A coupling was used on the motor shaft that allowed both the pulley and encoder for the motor controller to be used. This coupling caused a wobble in the motor shaft that translated into a variable speed in the drive shaft. The planned method of analyzing the vibration data depended on the speed of the drive shaft to be constant.
-Supporting the Shaft
The sleeve bearings used to support the shaft turned out to be unviable. The amount of friction created by those bearings was much higher than anticipated and the motor was unable to drive the shaft at any speed while the bearing was under load. In addition to this, the sleeve bearings were specified as press fitting onto the shaft, which in actuality was a slip fit. This would have caused excessive noise in the data.
Figure 9: Sleeve Bearings – Original Design
As a result, the original sleeve bearings were switched out with ball bearings, which allowed the motor to turn at the appropriate speed.
Figure 10: Ball Bearings – Alternate Design
Isolating the bearing from the motor with a drive belt worked as planned. The motor produces a lot of vibration while running. The drive shaft experiences almost none of that vibration due to the separation of from the belt.
Conclusions and recommendations
The test fixture is unusable in its current state. However, with simple modifications, the setup can be utilized in a manner, which will benefit the customer and their needs.
Since the band heater turned out to be a failure of a product, the team has researched possible alternatives. A cartridge heater was determined to be the most promising alternative. Holes should be drilled into the saddle to place the cartridge heaters to heat the entire saddle to the desired temperature.
Figure 11: CIR Series Cartridge Heater
As discussed previously, the coupling was not adequately supported. This is an easily addressable issue. A possible solution would be to add an additional support bearing onto the motor shaft. This bearing would not add any additional noise to the recorded signal due to the isolation created by the belt and pulley system.
The team did not have the time required to properly look into the bearings used to replace the support sleeve bearings. The bearings must be checked to ensure that they don’t have the same vibration response as the bearing being tested or an alternative to the ball bearing should be used.
References
[1]V.L. Parnell, S. Boedo, M.H. Kempski, K.B. Kochersberger, and M.H. Haselkorn Health Monitoring of LAV Planet Gear Bushings Using Vibration Signature Analysis Techniques
SAE 2007 Commercial Vehicle Engineering Congress and Exhibition Rosemont, IL, October 29-November 1, 2007
SAE Paper 2007-01-4190
[2]Shigley, J.E., Mischke, C.R., and Budynas, R.G., 2004,Mechanical Engineering Design. New York, NY: McGraw-Hill.
Acknowledgments
We would like to extend our appreciation to:
- The Xerox Corporation, specifically Melissa Monahan and Erwin Ruiz, for their financial support and use of their resources, both physical and intellectual.
- William Nowak, for his guidance and determination to get this project going and keep it going when times got tough.
- Dr Stephen Boedo, for the first-hand knowledge he provided, his assistance, and his interest in our project.
- Mitten Corporation, for their one-off pneumatic setup designed for this project.