1.0 INTRODUCTION

Throughout automotive, aerospace, biomedical, consumer, and numerous other industries, companies are striving to cut development costs, reduce time-to-market, and improve their competitive edge in today’s market. Since the mid 1970’s, computer-aided methods including CAE have increasingly played an important part in the product development process. Significant advancements in CAE technology allow product developers to visualize a concept and perform highly complex loading in a 3-D virtual environment. Using finite element techniques to analytically represent the part and then simulate load and constraint conditions, designers and engineers are able to predict the performance of a concept design in advance of any investment in tooling, prototyping, or physical test validation. ‘What-if’ studies can also be conducted to determine the effect of varying geometric features, material thickness, material properties, and/or alternative load cases. This allows manufacturers to have greater confidence that the best solution has been achieved before initiating a physical validation phase. Up-front CAE methods include optimizing the product for design features, weight, performance, and/or material. The benefits of using CAE methods early in the design process include shorter development cycles, reduced prototype and testing costs, and improved product attributes.

With the increased integration of simulation methods into product design has come the advancement of CAE technologies such as optimization and process automation. Optimization technologies allow the engineer to define a range of acceptable design parameters, and performance and mass targets and analytically iterate to identify the optimum combination of the variables to meet the objectives. Often times, tasks and processes of a repeatable, quantifiable nature (such as creating the analytical model, applying typical loads, generating electronic reports, and so on) can be easily automated using software, thereby improving the productivity and quality of the simulations.

The key to effective application of CAE lies in the ability to analytically represent the product and its environment including geometry, material properties, expected loads, and constraints. With consumer packaged goods, these requirements must also be balanced against more subjective criteria such as package aesthetics and volume efficiency. The following case studies demonstrate the process and benefits of applying an advanced analytical approach for the development of two consumer packaged goods - a Lever Fabergé bottle and a trigger spray bottle. Altair® HyperMesh® was used for finite element pre and post-processing in each of the studies. Altair® OptiStruct® was used for shape and size optimization of the trigger bottle, Altair® StudyWizard® for automating a DoE (design of experiments) and optimization study on the Lever pack, and additional commercially available finite element solvers were used for linear and nonlinear analyses.

2.0 CASE STUDIES IN PACKAGING APPLICATIONS

2.1 Lever Fabergé Bottle

Consumer product packaging designers are faced with conflicting requirements throughout the development process. Good package aesthetics are vital for the success of the product, while unit costs must be minimized and suitability for stacking and transportation maintained. A significant improvement in the design process can be gained if design information can be clearly communicated to the product designers early in the design process. This case study demonstrates how design optimization and advanced CAE can be used to deliver this. The resulting design process facilitates the early definition of an attractive package shape (Figure 1) incorporating features that will meet the structural and cost requirements. The design optimization process requires input in the form of a series of alternative shapes for the package, definition of a design objective (cost or weight) and constraints (structural, manufacturing). An automated series of structural assessments are then performed, and design sensitivity information and an optimum shape defined.

The design tool’s primary objective is to facilitate provision of clear information to the product design team about how to choose a shape which will be economical to produce, manufacturable, and capable of withstanding the required loads. It is extremely important that this information is generated in a timely manner. Automation of the process is therefore required wherever possible.

The design process is comprised of five major components. The first is the model generation process that includes importing CAD (computer-aided design) data, automated CAD cleanup and automated finite element meshing. The deliverable from this step is a baseline simulation model of the package as shown in the series of model generation pictures in Figure 2.

The next step, the parameterization process, uses advanced morphing technology to apply complex changes to 3D (3-dimensional) shapes, and thickness linking to allow weighted variation through the height of the bottle (size and shape variables are shown in Figure 3). The shape variables can be reviewed interactively and animated in 3D, making this complex information easily accessible to all members of the design team. The definition of the shape variables requires input from the design team to ensure that the look and feel of the pack is maintained and that none of the variables will cause manufacturing problems.

Size variables are defined based on knowledge of the manufacturing process. Control of the parison wall thickness in practice is limited to changing wall thickness in horizontal bands down the major axis. Design variables were therefore defined to parameterize wall thickness of the pack in achievable bands (Figure 4). At the end of the parameterization process, a series of design geometry variations are available for use in the optimization phase.

Figure 4: Definition of Size (Wall Thickness) Variables

The simulation environment for this application incorporates advanced solver technology to capture the non-linear collapse of the pack under the enveloping design condition (top load). This provides a means of understanding in a very short time frame (less than 1 hour) how the proposed geometry will perform, without the need for physical prototyping and testing. Before proceeding with the design optimization process, testing and verification of the simulation procedure on the baseline design is necessary. The top loading surface is moved vertically downwards to capture the peak buckling load and the post buckling behavior of the pack. Results from the simulation are presented in Figure 5.

The peak load capacity of the pack is influenced by the geometry, the redistribution of load due to contact with the loading platen and base plate and the non-linear material properties.

The collapse response under top loading (Figure 5) demonstrates that buckling first occurs in the neck region (figure on left). Load is then re-established before the base of the bottle buckles (figure on right). A comparison of the predicted response with actual test data revealed that the simulation environment accurately captures the real world response.

Figure 5: Stress contour plot of the baseline bottle during collapse simulation

ptimization is a two-stage process, which uses as input the baseline model and shape variables plus additional specification of the optimization objective and constraints. To start the process a Design of Experiments (DoE) study is performed. This yields a summary of the sensitivity of the design performance to the shape changes. This is followed by a full non-linear optimization to define the optimum shape. The whole process is set up and controlled from StudyWizard, which automates the procedure and simplifies user input.

The first step in the DoE study is a screening exercise to reduce the number of design variables. The design variables, which have least effect on the objective and constraints, are removed to leave the key design variables for the optimization phase.

The screening step identifies that the buckling capacity of the bottle is most influenced by the thickness at the top of the bottle. Significant sensitivity is also noted for global and local shape changes including: shoulder slope, footprint size and neck sculpting. A review of the data also indicated how the upper and lower bounds of the design variables affect the response. For example, increasing the lip thickness variable increases the buckling capacity, whereas increasing the sculpt depth at the neck has the opposite effect.

Seven design variables were selected from the DoE studies (main effects results shown in Figure 6) and were taken forward into the optimization phase.

Figure 6: Main Effects on Buckling Capacity from DoE Studies

The design optimization procedure finds an optimum combination of design variables to meet the objective (minimize mass) while satisfying the constraints (buckling capacity). The optimization generated a bottle design with the parameters summarized in Table 1. It is clearly demonstrated that design optimization can automatically provide the right mix of design parameters to save weight and increase performance.

2.2 Trigger Spray Bottle

The objective of this project was to demonstrate the application of virtual engineering tools to redesign a trigger bottle for a leading manufacturer of household cleaning products and achieve cost savings and improved structural performance. The project involved the integrated use of concept design methods and industrial design to meet three key objectives – increase the number of bottles/pallet, reduce packaging material, and reduce bottle mass. Application of optimization methods was employed to identify a better package design in less time than using traditional design approaches. Design and packaging criteria for the new design included:

Top load, bulge (fill load) and drop simulations were conducted on the current 32-ounce trigger bottle design to establish the baseline performance. Figure 7 shows the stress contour on the deflection profile of the current bottle under a top load alongside a graph of force versus time.

Figures 8 and 9 below show the stress contour and the deflection profile of the current bottle under a bulge load and drop load, respectively.

A footprint study was then conducted on five selected footprints with costant height but varying length and width. These footprints were rank ordered subjectively based on factors such as bottle stability, number of bottles per pallet, stack pattern and design efficiency. Concurrent to this study, concept sketches were developed for eight designs as shown in Figure 10.

These sketches were subjectively evaluated and rank ordered by a multi-disciplinary team based on factors such as volume efficiency, label area, aesthetics, grip shape, top load ability, and manufacturing. The top ranked design concept RADIUS was selected for development of a 3D model (see Figure 11), using the optimal footprint resulting from the rank ordered study. The model was then used to optimize the structural performance.

Shape and size (gauge) optimization was performed to optimize the bottle for top load resistance, bulge, and drop events while minimizing gram weight of the bottle. Figure 12 shows the location of the shape variables at the shoulder (shape variable 1) and at the wall fillet (shape variable 2), along with horizontal bands of wall thickness allowed to vary in the gauge optimization.

The optimized design detailed in Figure 13 resulted in passing bulge and drop test requirements with a 2% decrease in the weight of the (32 ounce capacity) bottle, and a 66% increase in the top load capacity as shown in Figure 14.

The use of virtual product development methods followed in this project allow for the development of a new, optimized bottle design, such as the bottle done for this project and illustrated in Figure 15, in approximately 4-6 weeks. In addition to faster time-to-market, substantial cost savings totalling over $2M in the first year of production have been realized as a result of several advantages of the new design:

3.0 Conclusions

A clear need and interest has been identified in the packaging industry for reliable design input early in the development process. The requirements for successful packaging design are many and conflicting, but the package must always remain attractive to the consumer.

The design technologies and process, applied to these two consumer goods package designs, can help bridge the gap between those involved in defining the right look and feel for the product and those involved in engineering the best solution. In both cases, the optimization results generated a reduced mass design concept given a baseline example design already on the market as a starting point. A first pass design optimization yielded a 2-5% reduction in bottle while exceeding the top-load capacity requirement and achieving further savings in manufacturing and transportation.

The high level of automation in the process facilitates rapid delivery of the design information. Advanced visualization tools and intuitive user interfaces make this information highly accessible to all of the design team.

The marketing team and product designers can get timely information about how to maintain or improve the appearance of the pack without compromising manufacturing or transportation performance. Careful choice of shape changes for the pack becomes a collaborative process between structural, manufacturing and product designers, with the design tool providing independent review. A by-product of the optimization process is readily accessible design sensitivity information, clearly indicating which shapes are beneficial to the pack performance. Extraction of the best geometry, which meets the requirements of all of the team is the final step in the process and the resulting CAD data can be taken forward for prototyping.

This design process provides potential for reducing design cycle times, through facilitating definition of strong design concepts early in the design process, which require fewer down-stream modifications. Close team collaboration is a requirement: design, marketing, manufacturing and engineering professionals are all called upon to review the proposed shape changes and understand their impact on the design.

4.0 References

The case study discussed herein on the Lever Fabergé bottle and the summary were derived from a detailed technical paper entitled ‘An Advanced Method for Optimizing Packaging Design’ by co-authors R. McNabb of Lever Fabergé, and Martin Kemp and Steve McMahon of Altair Engineering, Ltd and first presented in November 2002 at the Altair HyperWorks Conference 2002 in Coventry, UK.

Introductory portions of this paper, including the case study on the trigger bottle were first presented by the author at the Rapid-Pack 2003 conference sponsored by Schotland Business Research, Inc, on September 15, 2003 in Frankfurt, Germany in the paper titled ‘Simulation- Driven Product Design to Accelerate the Package Development Process.’