Team Berry:
Child Resistant Closure Liner System Phase 3 Design Report
Tuesday, December 12, 2006
Sponsor:
John Tauber
Advisor:
Michael Keefe
Mark Nauman
Andrew Seagraves
Jeff Staniszewski
Vincent Uathavikul
Table of Contents
List of Tables and Figures 3
Introduction 6
Work Accomplished 6
Phase 0 6
Phase 1 8
Phase 2 8
Initial Liner Concepts 9
Preliminary Testing 9
Metric Study: Cost Comparison 10
Concept Elimination 10
Phase 3 10
Concept Selection 10
Standard Leak Testing 11
Relaxation and Creep Testing 11
Finite Element Analysis 12
Test Design 13
Testing Tolerance Variation 13
Stress Limit Analysis 14
Detailed Design of the Prototype 15
Detailed Analysis of the Prototype 16
Statistical Analysis of Test Results 16
Implementation Plan 17
Conclusion 18
Appendix A: Cost Figures 19
Appendix B: Procedure for Standard Leak Test 21
Appendix C: Procedure for Elevated Temperature Leak Test 22
Appendix D: Concepts 23
Appendix E: FEA Data 24
Appendix F: Physical Testing Data 25
Appendix G: Wants and Metrics 26
Appendix H: UDesign 27
Appendix I: Gantt Chart 28
List of Tables and Figures
Figure 1: Flex Dome Closure…………………………………………...………….4
Figure 2: Original Instron Testing Rig………………………………….…...……..8
Figure 3: Leak Testing……………………………………………………....……..10
Figure 4: Sleeve (Yellow) and Punch (Green) Design……………..……..……….11
Figure 5: Torque Testing………………………………………………….…….…12
Figure 6: Instron Testing Rig……………………………………………..……….12
Figure 7: Compression Molded Liner and O-Ring Liner Models ...……………....13
Figure 8: ANSYS Model with Bottle Variation ...……...……….………….……..14
Figure 9: ANSYS Cap and Liner Model…..……………………..………………..14
Figure 10: Stress vs. Closing Torque…………………………….………………..15
Figure 11: TPE Liner Prototype…………………………………………………...16
Figure 12: Opening Torque Degradation Curve …………………....……………..17
Executive Summary
Over the last three decades, Berry Plastics Corporation has become the leading manufacturer of injection molded packaging in the United States. The company serves a broad range of major markets by providing hundreds of products in three product areas: open-top containers, consumer products, and most significant to our project, closures.
Berry is a significant supplier in a diverse closures market. They provide snap-on, threaded, dispensing, child-resistant and tamper-evident styles for food, dairy, personal care, household chemical paint and other market applications (http://www.berryplastics.com/company.html).
Aside from maintaining their standard product line, Berry produces numerous custom pieces for their customers. The proposed Senior Design project was to improve upon a failed child-resistant closure designed to seal without the aid of a liner. Before our team was introduced to Berry Plastics Corporation, one of their customers had requested a special child-resistant one piece closure aimed at reducing the relatively higher costs associated with manufacturing a standard 3-piece closure. Berry developed a prototype liner-less cap (Figure 1) that implemented a flexing dome to create a seal.
Figure 1: Flex-Dome Closure
Preliminary tests run by Berry’s research and development team showed that the closure would not be sufficient enough to maintain a seal for any desirable period of time. The failure of the Flex-Dome prototype prompted Berry, with the aid of our Senior Design team, to redesign the cap using a thin thermoplastic elastomer (TPE) lining in the dome of the closure.
When the project was first proposed to our team the goal was to design a TPE liner that would mold into the Flex-Dome cap and maintain a water-tight seal for a simulated period of one year through elevated temperature testing. Then, during early conferences between Berry executives and our Senior Design team, a correlation between closing torque and the pressure exerted on the cap was added to our project scope.
Benchmarking for this part of the design process was extremely hard since no other child-resistant closure on the market implemented our design. Instead of looking for similar closure designs, we examined the different materials and their properties that could be used in our device. The benchmarking verified that Berry’s initial desire to use a TPE liner and a polypropylene cap would be the best choices for our closure. We then compiled three designs that would satisfy Berry’s wants and proceeded in narrowing down the concepts to one final liner design.
We concluded that a liner of a constant thickness would best suit Berry’s needs and provide the most cost effective solution. To further verify our design, testing was performed using prototype bottles, caps, and liners. The results of the experiments proved that a liner with constant thickness in the contact area would create a watertight seal for an extended period of time. Also, Berry’s secondary want, to find a correlation between closing torque and pressure on the cap, was extracted from our test data.
The team determined that Berry’s initial Flex-Dome liner could still perform the task they desired with the implementation of a constant thickness liner. Furthermore, the cap and liner design would simplify the current child-resistant closures being produced by Berry Plastics. The reduction of parts would result in a shorter manufacturing time and ultimately save money for the company.
Introduction
Berry Plastics has asked us to design a liner for their Flex-Dome child-resistant closure. Initially, the cap was created as a liner-less sealing device, but the high probability of plastic deformation in the closure requires reevaluation. Berry immediately expressed interest in the liner being formed from thermoplastic elastomer (TPE), a product that familiar to their company. Also, the designed liner must be compression-moldable. Berry has told us that their goals for the project are to have the caps and liners create a seal to contain a liquid and maintain the seal for a year. The seal must have the ability to be duplicated on any bottle within a given tolerance range. It is also necessary to minimize the cost of production per unit to make this project valuable to Berry.
From our series of wants and constraints, an initial list of project metrics was determined using UDesign (See Appendix H) and then later refined after our leak test (See Appendix G). The seal performance can be tested and measured using a standard leak test and an elevated temperature test to simulate longer periods of time. These tests can be performed not only with bottles and caps at nominal sizes but also with combinations of the maximum bottle and minimum cap sizes and vice versa, in order to ensure performance in the entire range of tolerances. Evaluation of project costs show us that the main variable costs for the liner can be attributed to the volume of material used and the production cost of compression molding punches. The cost of the punch is a function of the machining costs involved in making a more complicated geometry.
While more TPE could be used for the liner to ensure a quality seal, it will cost more to produce. Similarly, a less complex liner profile will cost less to make, but may use a higher volume of material. Realizing that our main tradeoffs would be between the volume of the liner and the complexity of the punch, three concepts were created for consideration: one concept having the most TPE but the easiest punch to produce (constant thickness); another concept having a minimal amount of TPE and a complex punch (cutaway); and the third being in between the two extremes (variable thickness). We will be able to determine which concept is the best choice by weighing the economic and performance factors (See Appendix D for models of each design).
Work Accomplished
The final objective and scope of this project has evolved over the different phases. The tasks we initially identified during this process may or may not have been completed. As the project changed, priorities changed. In the following we will document the evolution of our project over the various phases, detailing the reasoning behind the path we chose.
Phase 0
When Berry Plastics presented us with this project, they had virtually decided upon the process (compression molding) and material (thermoplastic elastomer). Therefore, in Phase 0 we decided to verify this selection. We benchmarked the current processes and materials that they use as well as other options on the market. Also factored in was Berry’s current infrastructure and capabilities. In doing so, we found that the liner market was divided into the major categories of pulp-back, foam, and elastomer liners. The main processes to join the liner with the closure were simple insertion, two shot injection molding, and compression molding.
The metrics that we used to analyze the various material options were the ability to be shaped, elasticity, and cost. The pulp-back liners are some of the most common on the market and consist of a paper substrate with a metal foil seal. This material is inexpensive and widely used in their current manufacturing facilities. The main drawbacks are that it has virtually no elastic response once it is tightened, produces scrap material that cannot be recycled, and cannot be formed into complex shapes. Because of the profile of the closure we want to seal, we need to have a liner that can conform to the complex shape; therefore this material could not be used. Another liner material option is foam. This material comes in a variety of compositions, and can be inserted, injection molded, or compression molded. It has decent elastic properties and is moderately expensive, but foam can be crushed when exposed to forces for extended periods of time. We would want this material to seal for at least one year, and because the sealing quality would degrade over time, foam was not chosen. The Thermoplastic elastomer has very good elastic properties, can be manufactured by insertion, injection molding, or compression molding, and produces very little waste. The main drawback is that it is the most expensive of the three materials. It was chosen for its superior elastic qualities over time and its ability to be formed easily. Costs would be counteracted by reducing the material as much as possible.
As for the processes of putting the liner into the closure, the two main processes Berry currently uses are insertion and compression molding. Two shot injection molding is an option but its complexity and costs were high enough for it to be ruled out almost immediately. The determining factor in choosing the process was the ability to form the complex shape needed. Creating a complex shaped liner, inserting, and bonding it would prove to be time consuming and difficult. Compression molding would allow the material to conform to the closure quickly and with fewer steps; therefore we determined that compression molding is the best option for this project. Through benchmarking and material selection, we determined that we need to modify the original problem statement. The new problem statement is to create a compression molded TPE liner for the current closure.
Using the results from our benchmarking and metric development stages, we were able to develop target values for the liner. Our goals were to design a liner that could maintain a water tight seal during a two week leak test and an eight week elevated temperature test at 140ºF. We also felt that we could engineer the liner to work for all tolerance variations and then use finite element analysis to verify all of our results. We felt that completing all of our goals was an achievable task and could provide a validated and useful product to the Berry Corporation.
Phase 1
Using what we learned in Phase 0, we identified the two main objectives for this project: to design a liner and verify it through testing. To do this, we divided the processes into three critical subsystems. The first subsystem was the sealing pressure determination subsystem. The objective of this subsystem was to find a minimum torque that would seal the bottle. By finding the coefficient of friction of the materials involved, we would be able to analytically determine the pressure between the bottle land and closure. The next subsystem we identified was the finite element analysis subsystem. We would model the liner using the finite element program ANSYS and use it to determine which thickness of liner material would be the best to use in our system. The values of pressure that we would use in ANSYS would be taken from what the analysis found in the first subsystem. The final subsystem was the physical testing subsystem. The main purpose of this subsystem was to validate the analysis of the previous two subsystems as well as test the design we chose. In this subsystem we devised a way to collect data on pressure from an actual bottle. We also developed a protocol for leaking and creep testing.
The majority of Phase 1 was spent filling out all the specifics of each subsystem we identified. Drawings, theory, and models were developed in preparation for the testing that we would conduct in Phase 2. Some of the main assumptions that we made during this phase were that there was a threshold value for closing torque below which the bottles would not be able to be waterproof, and that the deflections would occur mostly in the liner. At the end of Phase 1, we had a clearly defined path of work that we needed to accomplish.
Phase 2
The major concern that we identified in Phase 1 was finding the minimum pressure that would seal a bottle. Since the bottles that we were going to use were still being manufactured, we had to focus on preparing tests that we felt could be useful in the future. We decided that we needed to build a testing apparatus that would work in conjunction with the Instron. This concept is illustrated in Figure 2.
Figure 2: Original Instron Testing Rig
Also, we wanted to perform finite element analysis in ANSYS that would model the cap and liner. These two steps would allow us to convert closing torque to a threshold pressure value and allow us to analyze the performance of different thickness liners.
Initial Liner Concepts
During Phase 2, we generated three concept liner designs. Realizing the tradeoff between the volume of the liner and the complexity of the liner, three differing concepts are easily compared. The first design is the simplest design, in which the stamp that would be created would have a flat profile, which we called the constant thickness concept. This would use the most amount of liner material but would be the least costly to machine the punches for compression molding. The next concept derived was the variable thickness concept. This would involve a constant thickness through the land region and a thinner thickness in the dome region of the bottle. The final concept we generated was a modification of the variable thickness liner, which we called the cutaway concept. This concept kept the design of the variable thickness liner but reduced the amount of material in the land region. The reasoning behind this would be for that there would only be a significant amount of material in the space where the bottle would contact the liner. Both the variable and cutaway concepts would require more machining of the compression molding punches, but would reduce the amount of material used. Pictures of the three concepts are included in Appendix D. In order to eliminate the concepts to determine which one we would pursue, we needed to compare them through testing and cost analysis.