Taking the Stress Out of Package Protection

Originally Published MDDI January 2003

PACKAGING

An examination of coupled predictive simulations of thermoforming processes and drop-impact loading in package design.

Balakrishna Haridas and Clinton A. Haynes

Thermoformed packaging is widely used today for the packaging of sterile medical devices and products. The benefits of thermoformed packaging are many. It offers easy control of shape and formability at an affordable cost, design flexibility for handling shock or impact, and an ability to maintain a sterile barrier. It also provides a diffusion pathway for EtO, gamma, and E-beam sterilization; a lightweight structure that reduces handling and shipping costs; and the delivery of brand identity for a device.

But recent medical device innovations—such as implants that incorporate onboard power systems, electronics, and delicate mechanisms for minimally invasive surgery applications—have put stringent structural requirements on packaging. These requirements usually pertain to shock and vibration protection from drop impact during handling or shipping.

The increased need for protection from such external loads is an added worry for manufacturers. To ensure that the appropriate structural-integrity criteria are met, the manufacturer must pay careful attention to the thermoformed package's structural design, material and gauge selection, and processing for optimal material distribution.

Thermoplastic resin systems such as polyethylene terephthalate-glycol (PETG), polyvinyl chloride (PVC), polybutylene terephthalate (PBT), and polyethylene naphthalate (PEN) are among the polymers employed for thermoformed medical packaging applications. In addition to package geometry and material distribution, other important parameters to consider in the design of thermoformed packaging in medical applications include resin material properties (e.g., modulus, yield strength, and ductility), poststerilization (EtO, gamma, or E-beam), as well as thermal stability (see Figure 1).

Amorphous and semicrystalline resin systems are known to exhibit marked changes in mechanical properties as a consequence of sterilization techniques.¹ Thus, while a package design and material distribution system may be adequate for sustaining structural loads such as drop impact prior to sterilization, the sterilization technique itself may compromise the package's mechanical properties and structural integrity.

For medical thermoformed packaging, drop testing has evolved as the critical test protocol for two reasons: assessing probable damage to the device, and evaluating the likelihood of a sterile-barrier breach due to package failure.

Although other loading conditions exist, such as compression due to stacking, industry has successfully shielded products from these hazards by placing the thermoformed package inside a secondary package. In turn, the secondary package is often further shielded by a corrugated shipping case. These expensive measures, however, are less effective when it comes to the rapid deceleration of the package and its contents during drop impact. In this extreme loading condition, the dynamics of the packaged instrument and the material characteristics of the thermoform can interact in a complex manner.

Currently, designing a thermoformed package is usually a reactive process. Feedback from package loading tests is available to the designer only after production-quality prototypes are available. The results of these tests and failed packages are the designer's only guide toward making design changes that satisfy both the structural requirements needed to protect the device and the need to minimize material usage (cost).

This empirical make-and-break process of structural discovery is detrimental to the product development cycle. That is especially true in situations in which the package development effort is initiated only toward the completion of the device. Make-and-break rework loops that are triggered as a consequence of failed prototype packages in drop tests can lead to costly project delays. Another major drawback of drop testing as the only technology for design assessment is that without extensive and often complex instrumentation of the package (e.g., accelerometers, strain gauges, high-speed video), it is usually very difficult to understand the fundamental physics that trigger failure modes—the key to driving design changes.

Nonlinear finite-element modeling has matured to the point that all stages of the thermoformed package development process can be accurately modeled and simulated.² The application of these methods, which range from forming simulation to assessing the interaction of the device with the thermoformed package under a wide range of loading situations, is now fairly routine. The designer's challenge of simultaneously optimizing a thermoformed package for both structural design and material usage can be completed without having to wait until production-quality packages are available for testing. This design-by-analysis process can be implemented early in the device development cycle, without the need for physical package samples. Implementing this approach guarantees that a robust package design is available by the time the device development and validation process is complete.

The Traditional Approach

In order to fully understand the opportunity that predictive structural analysis via finite-element methods offers to the package designer, it is useful to study the traditional thermoformed-package development process. Figure 2 describes the most common approach to developing a package for a medical device.

In most cases, the package design is subordinate to the development of the device. Therefore, by definition, it is not a focus until rather late in the device development process. Once the geometry of the device is established, however, the package design process begins, with the primary goal of accommodating the geometry of the product. Although structural integrity is a concern from the outset, without the benefit of predictive analysis tools the structural issues are dealt with using methods based primarily on experience and rules of thumb.

Once the initial package design is completed, a typically frantic tool-prototype-assemble-sterilize-test-redesign loop is begun. As time goes by and the launch schedule nears, the pace of this loop accelerates, generally churning out design modification after design modification, until one is identified that satisfies the minimum structural requirements—often at the expense of a heavier sheet gauge. This suboptimal thermoformed package design is then validated, submitted to FDA, and committed to production tooling. In efforts to accelerate the traditional process, it is not uncommon for manufacturers to skip the package sterilization step on the assumption that the material performance is more or less unchanged by sterilization.

The Predictive Engineering Approach

Although there is a general similarity between the traditional approach and the finite-element analysis (FEA)-based approach, the ways these development processes are executed are dramatically different (see Figure 3).

Once the first package design concept has been developed, it is prepared for virtual prototyping—rather than handing that design file to a tool builder to create a tool pass file (or to interpret the geometry from concept renderings or 2-D drawings). A combination of thermoforming process simulation and drop-impact structural analysis is used to identify the minimum sheet gauge, optimal thermoforming conditions, and package geometry (crumple zones, corner radii involved in impact, etc.) that will yield a material distribution that can reliably sustain drop-impact loads without failure.

Thermoforming Process and Drop-Impact Modeling

To illustrate the process of predictive analysis or virtual prototyping, we will examine a case study that used finite-element techniques to assess the response of an intraluminal stapler to a drop load. The drop simulations were conducted based on material-thickness distributions for the package; the material-thickness distributions were developed on the basis of a thermoforming simulation of the process. Thickness measurements and drop testing of physical samples were also completed and compared with analysis results to illustrate the level of accuracy expected from the simulations.

Thus, there are two key activities involved at this stage: the thermoforming process simulation and the drop-impact simulation.

Thermoforming Process Simulation. In this simulation, the package design concept is assessed for material distribution under manufacturing process conditions that would be employed to produce the package.

Plug-assisted thermoforming of medical device packaging involves preheating a polymer sheet (typically PETG) to just above the material's glass-transition temperature; then a plug is used in conjunction with vacuum or positive air pressure to form the sheet into the desired shape. Factors that affect the final outcome—i.e., the material distribution and end-use mechanical properties—include process variables such as the following:

Commercial technologies to conduct FEA simulations of thermoforming are now available to the engineer who is trained in the use of these tools. The behavior of polymeric materials at typical forming temperatures, however, poses interesting challenges. In the case of PETG for the stapler packaging application, the team involved in developing an accurate model would typically need to conduct biaxial stress-strain testing of the material at different temperatures. This would generate the required nonlinear stress-strain relationships required for input into the simulation. These relationships can be generated via specialized techniques such as bubble rheometry.

Figure 4 shows the finite-element models that were developed for the stapler packaging application. The plug-assisted forming process was simulated using the following process parameters and conditions:

Figure 5 shows the evolution of the wall thickness distribution and shape during forming that results in a final material distribution. The resulting correlations between measured data and predicted data are excellent. As the results in Figure 5 demonstrate, this modeling approach can also be used to assess the effects of initial gauge on the final outcome. Eventually, this predicted material distribution was mapped onto an FEA model using a 3-D interpolation algorithm. The FEA model was then used to conduct structural stress analysis of the package under drop-impact loading conditions.

Drop-Impact Simulation. Drop-impact simulation involves the modeling of the response of a packaged device to drop-impact loading. In this process, specific performance attributes of the package can be investigated and evaluated rapidly, without ever manufacturing and testing an actual package. These attributes may include the efficiency of the material distribution, the locations of high stresses or strains that cause failure of the package, as well as calculations of the magnitude of forces on the device as a result of the drop.

As in most development situations, the first design concept typically does not satisfy all the requirements, and more design iterations are needed. Using the quantitative information gained from the initial finite-element analysis as a guide, a design modification can be developed that addresses specific design deficiencies, and the analysis can be rerun. This process of physics-based design modification and reanalysis will continue and will rapidly converge at a specific design, because each iteration is based on a quantitative assessment of the previous concept.

During this virtual iteration phase, important design changes in package geometry, such as the addition of energy-absorbing features (crumple zones, corner radii), can be developed on the basis of drop simulations. This process contrasts dramatically with the traditional method, wherein a trip around the rework loop not only requires tooling modification (if the product can be reused), but also staffing and time to complete new process development and testing of physical units—a process that can consume weeks to months of schedule.

In order to simulate the interaction of the package with the device, it is necessary to have a structural model of the device included in the simulation. In these simulations, the device itself can be relatively crude. It only needs to represent the geometry and mass/stiffness characteristics, since the primary interest is in the packaging. Figure 6 shows a relatively crude model of a circular intraluminal stapler designed to fit into the thermoform. This approach ensures that the forces created by the device during the drop event are accurately transmitted to the thermoformed structure, and it does so without the computational baggage required to simultaneously solve the nonlinear structural equations for the structural performance of the device itself.

With the thermoform and device structural models complete in the case study, the appropriate velocity-boundary conditions were placed on the model. For drop-load simulations, the boundary conditions are limited to defining the desired orientation for the package as it makes contact. The thickness distribution resulting from the thermoform simulation was mapped onto the tray geometry for these impact models (see Figure 7). The entire package was rotated 10° (see Figure 6) to match the drop-test experiment (described below) and given an initial velocity of 4.429 m/sec toward the rigid plate. This is equivalent to a 1-m drop height.
 

Figure 8 depicts the transient evolution of impact-related stresses in the package during drop. From these frames it is evident that at the point of impact, the package deforms and continues to move over a certain distance before recoiling. During this period, the instrument inside the package continues to advance and decelerate as a result of its load and internal package features. The momentum of the device is transferred to the package and results in high stresses at these locations. This is in addition to high stresses and strains developed in the package in the region of impact.

Drop Testing. An instrumented drop test of the intraluminal stapler was conducted and documented using high-speed video. Accelerometers were placed on the stapler at two locations: one very close to the impact zone, and one further away. These transducers provided data on the G-forces developed in the device during the impact event.

Results of Drop Testing and Comparison with Simulations. A sequence of frames pulled from the video illustrated the severity of the drop event on both the package and the device. This confirmed the phenomena observed in the finite element simulations. From these frames it was evident that at the point of impact, the package deformed and continued to move over a certain distance before recoiling, as predicted.

Qualitative validation of the structural model can be achieved by comparing the video with the simulation results. The test results show that the maximum G-forces in the package during this 1-m drop test are in the range of 95 G. The analysis slightly overpredicts values in the range of about 100 G. With the exception of variation related to the differences in package orientation at the time of impact, it is clear that the test and analysis are in agreement.

Conclusion

Results from the case study show that FEA for thermoforming simulation provides excellent feedback on process-related parameters and material distribution. Accuracy up to 90% (wall thickness)—i.e., errors of 10%—can be achieved using such simulations. Successful results in the context of thermoforming process modeling is highly dependent on obtaining accurate data from the thermoforming manufacturer on tooling motions and process conditions. This processing model must then be closely coupled with structural analysis to assess whether the distribution of the thermoformed material will satisfy the structural requirements. Using this coupled approach, FEA technology for drop-impact analysis is easily leveraged in the hands of the trained engineer.

Accurate drop simulation provides a breadth of quantitative data that is simply not available via testing. For example, interactions between components of a device and package, and the distribution of G-forces inside the package, are directly available from FEA modeling results. In addition, the physics of kinetic energy absorption and dissipation via the interactions between stored elastic energy and viscoelastic and plastic deformation are only revealed via simulation techniques. Again, such energy absorption features as crumple zones can be evaluated using plots to aid the process of developing a package that is reliable under drop-impact conditions.

The predictive engineering approach is unique in that it represents a deliberate physics-based process that is driven by quantitative data. Key performance attributes, vulnerabilities, failure mechanisms, and sensitivities are known and understood before the package is produced. The product and package can be optimized before tooling commitments—i.e., the process of discovery—is eliminated. The advantages of this approach are realized in risk reduction that enables a product developer to manage project schedules, costs, time-to-market, and problematic rollouts that could result in franchise damage and even potential product recalls.

Today, prototyping and predictive engineering provide the technology and tools to predict the critical attributes associated with thermoformed package performance. This process does not eliminate the key activity of physical prototyping and testing; however, physical prototyping activities can be reserved for fine-tuning and validating successful designs that were developed using predictive simulation approaches.

REFERENCES

1. "Effects of Sterilization on Plastics and Elastomers," Plastics Design Library Handbook, 1994.

2. CA Haynes and B Haridas, "Reducing Cost, Development Time with Simulation, Analysis Tools," Packaging Technology and Engineering, 1999.

Balakrishna Haridas, PhD, is the director of SES Medical Device Technologies, a division of Stress Engineering Services (Mason, OH). Clinton A. Haynes is vice president of the company.

Copyright ©2003 Medical Device & Diagnostic Industry