Handle with care: designing damage-proof packaging for products.

While manufacturing firms focus on improving quality as a way to better compete in international markets, product-development teams are paying increasing attention to the packaging that protects those products during shipment. For no matter how good the product leaving the shipping dock is,

if it arrives at the customer's door in a damaged condition, repairs will be necessary and future sales could be compromised. This is not the only concern, however. Evershrinking product-development cycles mean less time to develop cost-effective protective packaging systems, so that speedy and efficient package design is becoming a higher priority.

Makers of valuable goods such as electronic equipment have traditionally taken a conservative approach to designing protective packaging, according to Alfred H. McKinlay, a packaging and handling technology consultant based in Pattersonville, N.Y. The resulting overprotection boosts the companies' packaging material costs and shipping costs, he said.

"Meanwhile, growing environmental concerns, particularly the stringent new solid-waste laws in Germany, have driven packaging designers toward what people in the industry call the three Rs': reduce, reuse, and recycle," said McKinlay, a former chariman of the ASTM Committee D-10 on Packaging.

These often-conflicting demands have led packaging specialists to develop a series of increasingly well-supported engineering principles and procedures that can rapidly optimize the protective function of a package so that the product will arrive safely for the minimum cost in money and environmental damage. At the same time, new technology such as improved laboratory shock- and vibration-testing equipment and miniaturized sensors designed to be shipped inside packages to measure transport hazards (including shock, vibration, temperature, and humidity) are being incorporated into these evolving package-development methodologies.

"This new equipment allows the packaging engineer to quickly evaluate the effect of changes in distribution of product/package systems, fine-tune package design while providing adequate package protection, evaluate the effect of manufacturing process changes on transportability, develop tests to make products more rugged, and speed up the product/package-development process," said Frank C. Bresk, vice president of Lansmont Corp. (Monterey, Calif.), a contract package-development laboratory and maker of shock- and vibration-testing equipment. Bresk is also a former chairman of ASME's Protective Packaging Committee.

Not too long ago, most packages were developed on a trial-and-error basis. Prototype product/package systems were evaluated by test shipping followed by disassembly and inspection. But because the shipping environment is not constant, many trial shipments were required, making the process costly and time-consuming. As a result, the quality of package development was based primarily on the experience level of the specialist.

This earlier situation is changing for the better. Many of the newer more-sophisticated packaging-development procedures are defined in consensus-developed standards issued by the ASTM, the U.S. Department of Defense, and other organizations. In 1983, ASME established the Protective Packaging Committee within its Electrical and Electronic Packaging Division to develop and disseminate sound engineering practices regarding the protective packaging of high-value electronic products.

In the meantime, more and more students are being taught these modern rationalized packaging methods at academic institutions that grant undergraduate and graduate degrees in packaging science. Among the major packaging schools in the United States are Michigan State University (East Lansing), Rochester Institute of Technology (RIT) in Rochester, N.Y., Clemson University (Clemson, S.C.), Rutgets University (New Brunswick, N.J.), Indiana State University (Terre Haute), and San Jose State University (San Jose, Calif.). Many of the students and professors at these schools are conducting advanced research that further improves package-engineering methodology.

Increasing numbers of these students are finding their way to manufacturing companies that have established in-house package-development facilities equipped with modern testing apparatus. The reduced shipping and packaging costs that result, as well as faster product-development times and quality improvements, are starting to convince managers of the worth of package development.

Unfortunately, "packaging is still in the transition from an art to a science," consultant McKinlay noted. "The technology and procedures are coming along a lot faster since I got into packaging, but there's a long way to go. Much of what we do is still based on what worked before."

Package Development

Designing an effective packaging system involves a step-by-step approach, McKinlay explained. In general, a packaging engineer needs three types of information to develop a protective-package system: data defining the damage-producing phenomena, mainly shock and vibration inputs, likely to be encountered in the shipping environment; the product's fragility characteristics and sensitivities to the expected environmental hazards; and the performance of available packaging materials.

The initial focus of the package-development process is on the product, Lansmont's Bresk said. This is because the engineer cannot optimize a package design without first knowing about the fragility of the product to be shipped. Determining the product fragility is accomplished by subjecting the unpackaged product to vibrations and shocks at incrementally increasing levels until damage occurs. Packaging engineers take the product fragility data, consider it in light of the expected distribution hazards, and design a protective package, working closely with the product designers to optimize the product/package combination and balance the cost of additional packaging and shipping fees against the cost of strengthening the product to withstand shipping forces. For example, if a component of a product repeatedly fails at a relatively low shock level, it may be more desirable to modify the components rather than develop an elaborate package, especially if it is early in the product design cycle.

Distribution Environment

The major hazards to shipped packages are shocks from handling drops, vibration from moving transport vehicles, and compression from stacking, according to Daniel L. Goodwin, chairman of the Department of Packaging Science at RIT. Other hazards include temperature and humidity extremes and stray electrostatic charges.

Packaging engineers typically employ general data on the distribution environments associated with truck, rail, and air transport that were compiled by organizations such as the ASTM and the International Standards Organization, Goodwin said. He noted, however, that each package in a shipment experiences a somewhat different shock and vibration regime due to variations in placement in the vehicle, different operators and routings, and accidental drops during handling. To develop data on a particular shipping environment, packaging specialists now send portable self-contained solid-state sensor/digital recording instruments through the distribution system to measure shock, vibration, temperature, and humidity. Two widely used examples of this type of instrument are the EDR-1 environment data recorder from Instrumented Sensor Technology Inc. (Lansing, Mich.), which contains three triaxially mounted piezoresistive accelerometers, and the DHR drop height recorder from Dallas Instruments Inc. (Dallas), which contains a piezoelectric triaxial accelerometer. The results of an investigation of distribution channel is a design drop height for shock and a frequency spectrum for vibration.

Shock environment is usually defined in terms of the drop height the package will experience based on the manner in which it will be handled, Goodwin said. "In general, the lighter a package, the higher a drop it can withstand. That is, the range of handling varies with respect to weight. A product that is handled by forklift will not receive the same treatment as one in a small parcel." Charts based on studies that define drop height as a function of size and weight of the package system indicate the probability of drops occurring from specified heights for a particular package weight. The packaging specialist selects a drop height that represents a given probability of shock input for that product.

Goodwin reported that researchers at RIT have recently been trying to reassess some of these standard drop-height profiles because of concern that they may be overly severe. In one case, technicians at Eastman Kodak Co. (Rochester, N.Y.) failed to find the amount of predicted damage to packaged rolls of Kodak printing paper stock following shipment through the' United Parcel Service (UPS) network, despite the fact that the packages did not pass the recommended ASTM drop-test procedure D-775, which specifies a sequence of free-fall drop tests.

"We sent a Dallas Instruments drop-height recorder to measure the handling events between Rochester and the UPS warehouse in Whittier, Calif.," Goodwin said. "The recorder uses two algorithms for deriving drop height. One takes into account the velocity change of the acceleration shock pulse that's measured and solves for drop height by integrating the area under that curve. The other algorithm uses zero-g methodology, in which the device records the time that elapses when the sensors find that they are in free-fall and calculates drop height from that." After interpreting the data, the RIT researchers "came up with a profile that matched the performance of Kodak's product/package system much more closely to the amount of dam- age we saw in the trial cases we sent to Whittier and back," Goodwin said.

The definition of the package's vibration environment is more difficult to determine and describe than its shock environment because of the complex and random nature of the input. The vibrational frequency spectrum measured during shipping is typically characterized in two ways. One method uses sinusoidal envelopes at the approximate acceleration levels that have been measured in shipments. The other approach, random vibration spectrum, duplicates shipping vibration more accurately. In this Case, the random signal is processed through a spectrum analyzer and converted into a power spectral density plot, which is expressed in terms of power levels or amplitude (g2/Hz) versus frequency (Hz). These characterizations are used later in the package-development process as programming inputs for laboratory vibration-testing machines. Random vibration does a much better job of re-creating vibrational damage because its random nature excites multiple modes in the product/package system.

Shock and vibration are not the only events on which distribution environment studies focus. Paul Singh, associate professor at the School of Packaging at Michigan State and current chairman of ASME's Protective Packaging Committee, told of another recent investigation that concentrated on compression. "We just finished developing a new data-acquisition system to measure the dynamic compression levels a package experiences in a mixed parcel load, such as a UPS containerized shipment," he said. "It's a load ceil-- basically a strain gage connected in a bridge circuit--the size of a small parcel." In tests that are soon to be conducted, the load cell will determine the maximum variation in compression that a small package will experience, providing data that will assist future package design.

Product Fragility

Shock sensitivities are reported in the form of a damage boundary curve, while vibrational sensitivities are shown by resonant frequency plots. The damage boundary technique is based on the shockresponse-spectrum (SRS) method of analyzing the effects of earthquakes on buildings, which was developed in the 1940s, said Lansmont's Bresk. Later, the advent of improved instrumentation and computers led engineers to apply the SRS method to the task of protecting high-value military hardware, but the approach was still too costly to use for optimizing packaging of commercial products. In the late 1960s, Robert E. Newton, now professor emeritus at the U.S. Naval Postgraduate School (Monterey, Calif.), considered means by which product shock fragility could be measured. Newton, who had used the SRS technique extensively for the U.S. Navy, developed a simplified envelope-type variation of the method for packaging, now called the damage boundary method. Tests showed .that the damage boundary method, which was based on a single-degree-of-freedom model, was accurate enough for use in packaging. In general, the approach was successful because it erred toward overprotection.

Somewhat later, Newton suggested the resonant search technique, now employed as a straightforward way to quantify product vibration fragility, Bresk noted.

Shock Fragility Assessment

The commonly used shock fragility test procedure is described by the ASTM D3332 standard. The object of these tests is to find the shock level at which the weakest component in the product fails.

In the procedure, the unpackaged product is rigidly mounted on the table of a shock-test machine--a device that guides the fall of an object and provides a precision shock impulse upon impact with the base. [Besides Lansmont, the other major U.S. producer of laboratory package-testing equipment is the L.A.B. division of Mechanical Technology Inc. (Skaneateles, N.Y.).] To avoid distorting the transmitted shock pulse, the attachment of the product to the table should be rigid. At the proper drop height it is released and allowed to flee-fall onto a shock programmer at the base, which controls the characteristics of the impact shock pulse. A rebound brake system prevents multiple bounces. Controlled parameters are acceleration, waveform, duration, and velocity change. The shock programmer controls are set to produce a low-velocity-change half-sine waveform shock pulse that lasts about 2 milliseconds. If no damage occurs, the procedure is repeated at slightly higher velocity changes until damage occurs. The last nonfailure shock input defines the critical velocity change for the product in that orientation.

Next, a new test specimen is mounted on the shock table. This time a high-pressure constant-force pneumatic shock programmer is mounted at the base of the shock machine. The programmer is set to produce a longer-duration (10-to-80-millisecond) trapezoidal waveform pulse with a low acceleration level and a velocity change about twice the critical velocity change that was determined previously. A similar testing procedure follows, in which the product is repeatedly sent into free-fall while acceleration levels are increased in small increments until damage results. As before, the last nonfailure shock input defines the critical acceleration for the product in that orientation.

These critical values are then plotted to produce the damage boundary for the product in that orientation. Ideally, damage boundary tests should be performed in both directions along each of the three orthogonal axes of the product. However, since this testing procedure is usually performed in the prototype stage of product development, few product samples are typically available. As a practical matter, limited numbers of samples are tested and several damage boundary routines are run on the same unit in different orientations.

The damage boundary plot for a product defines an area on a graph bounded by peak acceleration on the vertical axis and velocity change on the horizontal axis. Any shock pulse that can be plotted inside this boundary causes damage to the product. Acceleration is expressed in gs (multiples of the earth's gravitational constant), while velocity change, the change in a system's velocity magnitude and direction during a shock pulse, is the integral of the acceleration-versus-time pulse plot. Velocity change is proportional to the equivalent drop height (and the rigidity of the impact surface) through a simple kinematic relationship from basic physics.

Moro Accurate Tests

Though the previously described procedure is widely used today, packaging engineers are not standing pat with these test methods. Some of the assumptions inherent in the procedure are coming under reexamination as part of the regular standards-review process. "In shock testing, the impulse is of short duration because we're assuming that the product is heavy enough (low-frequency) that it's reacting to the impulse, not the acceleration," said Terry Baird, a quality engineer at the Disk Memory division of Hewlett-Packard Co. (Boise, Idaho). "But our products-- hard disk mechanisms--are always getting smaller, lighter, and stiffer, so for the new products, the test impulse may no longer be short enough to ensure that we're measuring the effect of the impulse rather than that of the acceleration. We're now considering how we might change this approach.

"Similarly, when we impart a trapezoidal waveform to a product, we're ignoring the distortion caused by the vibrational modes of the table itself," Baird said. "But it may be that this overlaid high-frequency oscillation or noise may be causing the damage we see in some product components. This type of subtle failure mode is tough to separate out. It's likely you'll find engineers buying 'cleaner' pulsers that are more specific to the application in the future."

Vibration-Fragility Assessment

Vibration damage is most likely to occur when some portion of the product is excited into resonance, which can cause fatigue failure, Bresk said. In general, products will not be damaged by nonresonant inertial loading caused by vibration in the distribution environment because the acceleration levels of most vehicles are low compared to a product's critical acceleration. The goal of vibration fragility testing is to identify the product's resonant modes in each of the principal axes. This is done by vibrating the unpackaged product at a constant acceleration level while varying the frequency of vibration according to ASTM D3580 recommendations.

To perform a resonant frequency search test, the product is mounted on the table of an electrohydraulic vibration system and subjected to a constant acceleration input at a low level (typically 0.25 to 0.5 g) over a range of 2 to 300 Hz. Most resonant search tests are run in the vertical direction because the most severe vibratory inputs occur in this direction, Bresk said. Accelerometers are fastened to critical components within the product to determine their response to the input acceleration. The ratio of the response to the vibrational input (the "transmissibility") is plotted against frequency. This ratio reaches a maximum at the resonant or natural frequency of the component.

Protective Package Design

Armed with distribution environment data and the product's fragility characteristics, the engineer is now in a position to design a protective package. In addition to shock- and vibration-resistance data, the packaging engineer considers several other factors, including package size, shape, materials, compressive strength, humidity expected to be encountered, protective coatings, closure methods and their effect on integrity, protection of product surfaces and labels, protection from electrostatic discharge, provision for shipping accessories in the same package, recycling, whether the container is reusable, whether the contents are load-hearing, and whether the product is to be shipped alone or palletized.

Electronic equipment is typically shipped in boxes of corrugated paperboard. Package test standards--including compression, drop, horizontal-impact, vibration, and severe climate tests--have been developed by the ASTM Committee D-10 on Packaging and other organizations.

A keytask at this point, according to McKinlay, is to evaluate and choose the cushioning material, optimizing for efficacy versus cost. A cushion is a mechanical isolator that mitigates the effect of shock and vibration on a product by deforming in response to induced forces. Cushions transform higher-g short-duration shock pulses caused by impacts into lower-g longer-duration pulses. A wide variety of cushioning materials is available, including polystyrenes, polyethylenes, and polyurethanes.

It is crucial for the package designer to know the dynamic characteristics of the candidate cushion materials. Basic cushion-performance data are available from the material suppliers, but often the packaging engineer must develop specific performance data related to a particular cushion configuration.

Two related cushion-performance plots are used. One describes the shock-absorbing ability measured by accelerometer-instrumented impacts conducted in accordance with ASTM D1596, comparing static stress that the product's weight imposes on the cushion (weight divided by support area) and peak acceleration due to impact. These shock curves show the peak acceleration that will be transmitted by various thicknesses of the cushion for different values of static stress.

The other plot, called an amplification/attenuation curve, describes the material's vibration transmissibility response. It is developed by resonance search tests that involve subjecting spring/mass systems consisting of the cushion materials and various test weights to vibration inputs over the frequency range typical of the distribution environment and measuring the responses. The frequency plot.

Other pertinent information about the cushion materials include compressive stress versus strain properties (which indicate the degree of linearity and cushioning efficiency), compressive creep characteristics (loss of thickness when subjected to a long-term compressive load), and compressive set (a measure of the permanent loss in thickness after the removal of a static load).

In choosing the optimal cushion material, the designer looks at the shock curves for the design drop height, determining which material, thickness, and configuration will limit peak acceleration to the critical acceleration value. Then using the static stress determined from the shock curve, he or she uses the amplification/attenuation curve to find whether the cushion will attenuate the critical resonances. The cushion material must satisfy both shock and vibration criteria simultaneously.

Cost, of course, is also a big consideration. Normally, several materials fit the bill, and the designer picks the least expensive.

Product/Package Testing

Once the design has been finalized and a prototype fabricated, the product/package system is tested using inputs that duplicate the potentially damaging effects of the distribution environment to verify its performance, Bresk said. It is necessary that a product/package system be subjected to the design drop height and the chosen vibration spectra, because the previous development prosumptions that can only be evaluated by test. For example, the cushions have a different shape than test samples, and while real-world inputs occur in three dimensions, the individual lab tests were conducted alone in a single axis.

These procedures are covered by ASTM D4169, which attempts to correlate lab tests with field experience. Tests are arranged in the same order as the hazards that occur during shipment, Bresk noted. For example, to simulate transport effects on a package to be shipped unpalletized in a motor truck, the package is first conditioned in a temperature and humidity chamber and then subjected to drop tests to simulate the manual handling events that may occur when being positioned in the truck. Next. the package is tested in compression, replicating the effect of stacking in the trailer. Afterward, it is placed on a vibration table that is programmed to perform random vibration tests. Finally, drop tests are repeated to simulate unloading at the end of the journey.

Using these modern consensus-developed procedures, package development has moved away from trial and error and toward rationalized engineering practice. In the process, product-development engineers are finding that packaging engineers can be a valuable part of the product team. Fragility data obtained in the package-development process often shorten the product-development cycle by pinpointing problems with product integrity earlier in the design cycle than might occur otherwise.