Yam, Kit L. (ed.). The Wiley encyclopedia of packaging technology

Подождите немного. Документ загружается.

value at sea level and is, therefore, easily compared from

one laboratory to another.

Equilibration. A sample of barrier ﬁlm is mounted in a

permeation instrument test cell and the test is started.

Note the time data in Figure 1. The results of a test are

valid only once the barrier material has reached

equilibrium.

. Material thickness

. Relative humidity

. Temperature

. Time

. Barometric pressure

The time for the sample to reach equilibrium cannot be

appreciably accelerated or shortened. Time is a physical

function of the material. The better the barrier, the longer

it takes to get results. For example, 1 mil (25 mm) PET

takes 1 h, whereas 5 mils (125 mm) PET takes 8 h to

stabilize. If the result is taken prematurely, it will be

wrong, as steady state has not yet been reached.

Five major factors that affect permeation rate through

a barrier polymer are listed in Figure 2.

PIONEERING METHODS

We must acknowledge some of the pioneer work in the ﬁeld

of permeation, in particular the water-vapor permeation

cup test (1, 2) (see also Figure 3) developed in the early

1940s. These test methods determine the loss of water from

a cup through a barrier sealed on top of the cup and are still

used in a few locations. Permeation does have disadvan-

tages, such as a long time to obtain results (days vs. hours),

limited testing range, being labor intensive, requiring

expensive climate-control chambers, and being operator

dependent. However, steady progress has been made in

instrumented testing systems that allow shorter test times,

greater limits, better repeatability, increased sensitivity,

computerized control, and documentation.

Also, recognition is due to other early developments as

listed in ASTM D1434 (M) on the manometric cell (also

known as the Dow cell) and ASTM D1434 (V) on the

volumetric cell. These were gallant attempts, but are no

longer in widespread use.

CURRENT METHODS

Oxygen Transmission Rate (O

TR). Oxygen is often re-

ferred to as the ‘‘thief’’ of ﬂavor, texture, color, nutrition,

and shelf life. The importance of testing O

TR was recog-

nized early in the development of barrier materials such

as coated papers and plastics. The coulometric sensor (also

known as the coulox sensor) soon dominated because it

could detect O

down to the parts per billion (ppb) level,

was stable, and was easy to maintain and reliable.

Initially only ‘‘dry’’ tests were performed, 0% rh, which

was adequate for many of the barriers of that day. How-

ever, subsequent barrier materials were developed that

had superior O

permeation characteristics, but the O

of these new materials was a function of humidity.

Oxygen permeation testing has advanced in capability

to allow testing of the most recent developments in barrier

materials. The ability to control the percent relative

humidity (% rh) independently on each side of the barrier

has been one of the most signiﬁcant advancements. This

allows a better simulation of true-to-life shelf conditions

and, therefore, more credible results.

The effect of % rh and temperature on O

TR can be

observed in Figures 4 and 5.

Performance expectations are important. The measur-

able range of O

transmission rates is 0.001 mL/(m

day)

to over 1,000,000 mL/(m

day). The following list is spe-

ciﬁcations on Ox-Tran systems,

as used in ASTM D3985

and F1307:

Sensitivity—ﬁlms, 0.01 mL/(m

day) [0.0006 mL/

(100 in.

day)]; packages, 0.00005 mL/package day)

Figure 1. Typical example of time to condition specimen.

Figure 2. Five major factors affecting permeation rate.

Figure 3. Wet cup test.

1208 TESTING, PERMEATION AND LEAKAGE

Repeatability—ﬁlms, 70.05 mL/(m

day) [0.003 mL/

(100 in.

day)] or 1% of reading; packages,

70.00025 mL/(package day) or 1% of reading

A new high-sensitivity method and apparatus now

makes it possible to attain the following speciﬁcation:

Sensitivity—ﬁlms, 0.001 mL/(m

day) [0.00006 mL/

100 in.

day)]; packages, 0.000005 mL/(package  day)

Repeatability—ﬁlms 70.005 mL/(m

day) [0.0003 mL/

(100 in.

day)] or 1% of reading; packages,

70.000025 mL/(package  day) or 1% of reading

The importance of temperature control cannot be over

emphasized. Relative humidity is a function of tempera-

ture, and to control % rh is not an easy task because it can

vary by 5% rh/ 1 C in a region of the test.

Measuring High O

TR Values. At ﬁrst, all efforts were

directed toward low permeation readings (high barriers).

An emerging need also exists to measure high O

values. Initially, the use of a mask (to decrease the

permeation test-cell area) was used for years. As higher

TR were required, new test techniques were developed

to yield more repeatable results and to keep the O

sensor

from swamping. With care, values of r155,000 mL/

day) can be measured. So the measurable range of

TR is 0.001 mL/(m

day) to over 155,000 mL/(m

day).

The following international test methods, as shown in

Figure 6, were written around the use of the coulox sensor

for oxygen and the infrared method for water vapor, which

makes it possible for laboratories in different parts of the

world to compare O

TR and water vapor transmission rate

(WVTR) test results.

Water-Vapor Transmission Rate. After the cup test, in-

strumented systems (ASTM F372, F378, TAPPI 378), and

infrared (IR) systems were developed that achieved re-

sults in a shorter time (hours vs. days). Comparison of

WVTR using the IR method and the cup test revealed

remarkably similar results. Today, it is estimated that

95% of WVTR testing in the United States is done using

ASTM method F1249 and TAPPI method T557 pm-95.

It was standard practice, at one time, to run all WVTR

tests at a given rh, (e.g., 100% rh). Then, if the WVTR at

another % rh was desired, then one would simply multiply

the 100% rh results by the % rh of interest. For example,

the 100% rh result was multiplied by 0.8 if a result at 80%

rh was of interest. This was done because there was no

convenient, clean method to generate precise % rh values.

The results using this technique, called factoring, can

be observed in Figure 7.

Notice the excellent correlation between these two

methods of creating the driving-force RH. The test results

from four common packaging ﬁlms were almost identical,

no matter whether the actual rh was created with salts or

distilled water, and the result then factored.

Considerable advantages have been obtained with

factoring, including

1. Elimination of messy, uncontrollable salt solutions

2. Greatly reduced possibility of instrument corrosion

3. Increased ease of equipment operation

This method of testing without salts is widespread today.

By about 1990, however, it became apparent that this

technique may not work for all barrier materials, most

notably hydrophilic polymers. In these cases, water vapor

is so soluble that the ‘‘factoring’’ relationship does not

Figure 4. Effect of relative humidity on O

transmission rates of

several plastic ﬁlms.

Figure 5. Effect of temperature on O

transmission rate of 0% rh

for several plastic ﬁlms.

Figure 6. International standards for permeation and transmis-

sion-rate testing.

TESTING, PERMEATION AND LEAKAGE 1209

seem to hold true. Nylon 6 is a good example of this (see

Figure 8).

As can be observed, factoring does not work for all

barriers. Therefore, a system was developed that is more

universally applicable by creating humidity and tempera-

ture test conditions close to actual shelf conditions of the

package. The advantages are as follows:

1. Tests can be conducted close to actual shelf conditions.

2. A range of RH values is available.

3. Relative-humidity values easily changed.

4. No mess.

The water-vapor transmission rate is expressed in g/

day), or g/(100 in.

day). The measurable range of

WVTR is 0.0001 g/(m

day) to 1000 g/(m

day). The spe-

ciﬁcations of some WVTR systems are as follows:

Sensitivity—ﬁlms, 0.0001 g/(m

day) [0.000006 g/

(100 in.

day)]; packages, 0.000001 g/(package day)

Repeatability—ﬁlms, 70.002 g/(m

day) [70.00015 g/

(100 in.

day) or 1% of reading; packages, 70.00

001 g/ (package day) or 1% of reading

Carbon Dioxide Transmission Rate (CO

TR). An IR sen-

sor is used to detect CO

, and when incorporated in an

instrument system to detect permeation, it yields CO

TR.

The major applications are for cola drinks (CO

escaping

from PET containers), cheese packaging, and other CO

sensitive products. There is no way to extrapolate the

action of one gas to determine the action of another gas.

Each gas is independent of the other gases and must be

tested as an individual gas to determine its TR.

The measurable range of CO

transmission rate is

1.0 mL/(m

day) up to 100,000 mL/(m

day). Speciﬁca-

tions for CO

systems are

Sensitivity—ﬁlms, 1 mL/(m

day) [0.07 mL/(100 in.



day)]; packages, 0.005 mL/(package day)

Repeatability—ﬁlms 75 mL/(m

day) [7 0.35 mL/

(100 in.

day)] or 1% of reading; packages,

70.025 mL/(package day) or 1% of reading

TR permeation systems can detect 100 ppm CO

This is an important factor when considering permeation

or leakage. Efforts have to be made to keep the CO

from

being diluted during the testing process. Diluting the CO

level can make a bad situation look good. This is just the

opposite for O

. Diluting a low-O

signal with air (20.8%

) will make it worse, not better.

An outstanding example of partial pressure can be

observed using CO

. Fill a PET container with 100%

and seal securely. The partial-pressure phenomenon

happens over a period of time. CO

leaves the PET

container much faster than O

or N

air enters the

container. As CO

leaves, a vacuum is formed, which

eventually causes the container to collapse. Even though

was removed, it left a vacuum, and O

and N

followed the partial-pressure law. The vacuum seems to

play no part in the permeation process in this case.

Aromas and Flavors. As stated above, these need to be

tested as individual gases, and there are a host of these in

Figure 7. Comparison of four ﬁlms ‘‘factored’’

with 100% rh vs tests with a saturated salt solution

of zinc sulfate. (Saran Wrap is a registered trade-

mark of Dow Chemical USA.)

Figure 8. Comparison of nylon 6 ‘‘factored’’ with

100% rh vs tested with a saturated salt solution.

1210 TESTING, PERMEATION AND LEAKAGE

organics. This is a developing ﬁeld with many opportu-

nities, such as citrus ﬂavor,

D-limonene from an orange-

juice container, citrus ﬂavor into other food products (e.g.,

breakfast cereals), petroleum gases out of packages and

into other packaged goods, and so on. Fortunately, a

technique has been developed to handle all these organic

molecules, however, only one at a time.

In brief (referring to the Aromatran schematic in

Figure 9) a carrier gas (helium, N

, or any number of

inert gases) carries the test gas (

D-limonene, butane,

methane, apple ﬂavor, etc.), which has permeated through

a barrier to a cold trap, where it is captured and accumu-

lated over a period of time. At the end of the ‘‘accumulate’’

time, the cold trap is heated quickly, which thereby

releases the test gas molecules that had been captured.

The carrier gas, which is not affected by the cold trap,

continues to ﬂow and carries the test-gas molecules to a

sensor. The sensor is an FID that is set up to detect the

amount of test gas captured.

The results are given in common permeation terms:

X TR ¼ mL=ðm

 dayÞ at 30



Permeation becomes more critical as the shelf life is

extended. Not only is there concern about gases, such as

, water vapor, and CO

entering and leaving packages,

but also the effects of aromas and ﬂavors are playing an

important role in the industry.

LEAKAGE

Everyone knows what a leak is—leaks in the roof, a leak in

a tire, even a secret can be leaked out. It is something

everyone can deﬁne, and everyone has an opinion on the

subject.

In the previous section, we discussed permeation. Both

permeation and leakage do the same thing: shorten shelf

life.

Permeation can be determined, as described above;

however, leaks are more erratic, evasive, and unpredict-

able. Where a permeation value might hold true for a

production run of plastic ﬁlm, leaks can vary from one

package to the next. The ideal is to have 100% leak testing

on a production run. This is ideal but not practical in most

applications.

The perpetual issue of ‘‘hole size’’ is debated among

manufacturers of food, pharmaceutical, and sterile medi-

cal products. Each has different criteria. The hole that will

allow bacteria to enter a sterile package is much smaller

than a hole that will allow gravy to drip out of a microwave

entre

e. First, we must agree on a term called ‘‘effective

hole size.’’ This is the summation of all the holes in the

package. It could be the summation of round holes, square

holes, holes with long tunnels, slight tears, big tears,

hairline cracks, and so on. Let one hole represent the

combination of all the leaks. By so doing, we provide a way

of comparing one package to the next. Figure 10 illus-

trates the ﬂow rate that one would expect through a

round-style hole under various differential pressures

across the hole. These curves apply to air at ambient

temperature through a hole in a thin material.

DESTRUCTIVE TESTING FOR LEAKS

Waterbath or dye–leak test. (See Figure 11). Throughout

most of the twentieth century, package integrity has been

veriﬁed by placing the package under water and looking

for bubbles. Variations of the ‘‘waterbath’’ test have gen-

erally involved attempts to improve on the problems

inherent in looking for bubbles. The ‘‘dye test’’ was an

Figure 9. Aromatran schematic.

Figure 10. Flow rate of air through a round hole.

Figure 11. Waterbath or dye leak test.

TESTING, PERMEATION AND LEAKAGE 1211

improvement, but the two remaining problems are lack of

repeatability and destruction of the package. Advantages

and disadvantages of these tests are as follows:

1. The waterbath test

a. Advantages

(1) Can provide sensitivity

(2) Inexpensive equipment requirements

(3) Low-level technology, easy to train operator

b. Disadvantages

(1) Destructive

(2) Results are very operator-dependent

(3) Expensive labor requirements

(4) Messy

(5) Capillary attraction can give false results

2. The dye test

a. Advantages

(1) Ability to test many samples at once

(2) Inexpensive equipment requirements

(3) Low-level technology, easy to train operator

b. Disadvantages

(1) Destructive

(2) Unreliable and operator dependent

(3) Slow and labor intensive

(4) Messy

(5) One-way valve effect

Pressure/Pressure-Decay Method. This technique allows

two tests to be run sequentially: (a) burst/seal strength

and (b) leak.

Burst/Seal Strength. This is a practical starting point. It

is important to ensure that handling and shipping does

not destroy the package, which left the plant without

leaks. It must remain sealed to maintain adequate shelf

life and/or the desired quality when it reaches the con-

sumer. Air pressure is introduced into the package until it

bursts! The burst-point pressure is held so that it can be

recorded. This burst test should be repeated several times

to obtain an average. Analysis of this data would be

required if damaged packages were discovered by a con-

sumer. This also allows research and development (R&D)

to investigate new packaging materials and sealing tech-

niques. The average test time is 15 s.

Leakage. Once the burst/seal strength has been deter-

mined, look for leaks. Introduce a pressure into a new

package at less than 50% burst pressure and monitor

pressure decay. Pressure decay is a function of effective

hole size and pressure differential and is inversely propor-

tional to the volume of the package under test. It has been

shown that holes as small as 1-mil diameter are detectable

in small packages such as pharmaceutical bubbles, but it

would be almost impossible or impractical to sense a 1-mil

hole in a blimp. This method does not indicate the location

of leak(s). The average test time is 10 s.

Other techniques that are destructive include halogen

detection where a halogen agent, such as iodine, bromine,

or ﬂuorine, is used as a tracer gas. The system is pressur-

ized with the tracer gas, and a probe is used to detect a

leakage point. Although this test does detect leakage

points, it is highly operator dependent; thus, a poor

operator will get the best results, only on the basis of

fewer rejects. The instrument response is slow and lacks

quantitative accuracy. Food or drug products are not

normally tested by this technique.

Halide Torch. Similar to the halogen system described

above, in the halide-torch test, the tracer gas is sucked

into a burner through a hose attached to the bottom of a

burner. A change in the color of the ﬂame indicates a leak.

Thermal-conductivity detector. A tracer gas, such as

helium, carbon dioxide, or butane, is pressurized in the

package, and a probe senses a change in gas thermocon-

ductivity as the tracer gas enters the probe. This

method does detect the leakage point; however, it also is

highly operator dependent because if the probe is moved

too fast or too far away from a hole, nothing will be

indicated.

Helium Leak Detector. This has the greatest sensitivity.

Inﬂate the package with helium. Then, using a probe, sniff

around the package to ﬁnd a leak. The major disadvan-

tages are, besides high operator dependence, initial costs

(for supplies, etc.) and unit maintenance, that if the

instrument is swamped, it takes hours or days for a skilled

technician to get it back on line.

Headspace analysis. A practical way to check for leaks is

to measure the headspace of a package, and this can be

done at any time, such as immediately following sealing or

later, or even years later. Using a syringe, a sample of the

headspace is taken from the package and injected into an

instrument. If the O

content is to be determined, a

zirconium oxide sensor is used; if CO

content is to be

determined, an IR sensor is used. The results are given in

percentage of contents, such as 0.5% O

or 30% CO

, and

within a few seconds after injection of the sample into the

instrument. Applications include checking for cap/map

effectiveness, shortly after testing to determine whether

sealing was effective and met speciﬁcations and much

later, to determine whether permeation had any effect.

Note, in this latter case, that the long-term effect could be

either a slow leak or permeation; this instrument cannot

separate the two.

Nondestructive Testing for Leaks. The CO

leak detector

is most commonly used in nondestructive testing of food

and pharmaceutical packages. It is ideal if the package

has been ﬂushed with CO

before sealing or if CO

generated inside, as in the case of coffee.

Leak Detection Using CO

Tracer Gas. Figure 12 shows

the ﬂow diagram of a CO

leak detector. The pump

pictured is always running, sweeping room air in from

the left and purging the test ﬁxture and CO

sensor. When

the test sequence is initiated, the dwell solenoid closes,

thus causing the pump to draw a vacuum on the package

inside the test ﬁxture, which draws CO

out of the leaks.

At the end of the test cycle, room air is again allowed to

1212 TESTING, PERMEATION AND LEAKAGE

sweep into the test ﬁxture. This ﬂushes any CO

gas into

the CO

sensor. If this CO

concentration exceeds the

alarm threshold, the instrument signals that a leak is

present. Throughout this cycle, room CO

concentration is

viewed as the unit’s zero, i.e., reference, point.

An important factor is the size of the test ﬁxture

compared with the package under test. The sensitivity is

inversely proportional to the space outside the sample

being tested. As mentioned, care must be taken to avoid

diluting the CO

signal. Therefore, a specially designed

ﬁxture is needed for best results, i.e., to sense small leaks,

and in most foods and/or pharmaceutical applications,

small leaks are important.

In those cases where no CO

is in the package, CO

can

be forced into the packages through the leaks by use of a

pressure chamber ﬁlled with CO

. The vacuum, shown in

Figure 14 then draws CO

out of the package, which is

sent to the CO

sensor. The tests that use this technique

take 15 s on average.

Carbon dioxide trace-gas leak detection has been

shown to be a sensitive, repeatable, and nondestructive

alternative to traditional waterbath and dye-leak techni-

ques. Justiﬁcation is easiest when the product is expen-

sive, as this nondestructive test procedure enables product

to be exposed to additional laboratory tests, or to be

returned to the production line.

CONCLUSION

The search for better leak-testing methods continues. The

challenge is to devise a system that is nondestructive, can

conduct 100% inspection at high speed, and, of course, is low

cost. To date, the diversity of applications has precluded this.

Each application must be evaluated to determine the best

solution. Testing-systems manufacturers are challenged to

write their speciﬁcation legibly, clearly, and precisely to

ensure that they are understood and can be applied.

Permeation is well deﬁned and accepted in American

Society for Testing and Materials (ASTM) speciﬁcations.

Leakage requires the same attention.

BIBLIOGRAPHY

1. ASTM E96, American Society for Testing and Materials,

Philadelphia, PA.

2. TAPPI 464, Technical Association of the Pulp and Paper

Industry, Atlanta, GA.

TESTING, PRODUCT FRAGILITY

H. H. S CHUENEMAN

San Jose State University,

San Jose, California

The vast majority of products undergo a certain amount of

handling and transportation from the time they are

manufactured until they are ultimately used. This is

especially true in a global economy. Normally, products

are enclosed in a protective-package system during the

time they are exposed to the hazards of the distribution

environment. This article outlines engineering principles

and procedures that optimize the protective function of a

package system and help guarantee the product’s safe

arrival at favorable cost.

The waste of resources caused by improper packaging

is enormous. Whether this waste shows up as damaged

product or an overly expensive package system, it is still

waste that can be avoided. The application of sound

engineering principles to optimize the product protection

system must receive signiﬁcant endorsement by product

managers, quality assurance (see Speciﬁcations and qual-

ity assurance), and all those concerned with delivering a

quality product to the customer.

To develop a protective-package system, three impor-

tant pieces of data are necessary: (i) information on the

likely hazards (shock and vibration) from the distribution

environment; (ii) product fragility characteristics and

sensitivities in terms of the environmental inputs; and

(iii) the performance characteristics of commonly avail-

able packaging materials.

The engineering process for developing a protective-

package system involves a step-by-step problem-solving

approach similar to other engineering disciplines. The

principal focus is on product characteristics because it is

impossible to optimize a package system without ﬁrst

knowing basic engineering data of the product itself.

Once the environmental hazzards have been deﬁned,

and product fragility has been determined, the engineer

can evaluate the economic feasibility of possible product

modiﬁcations. For example, if a certain component within

the product keeps failing at a relatively low input level, it

may be economically more desirable to modify that com-

ponent and increase its ruggedness rather than design an

expensive and inefﬁcient package system for the entire

product based on the sensitivity of that one component. If

it is economically feasible to modify the product in order to

effect a package-cost savings, this option should be studied

carefully. Many examples exist where slight product

Figure 12. Co

trace-gas leak detection; nondestructive.

TESTING, PRODUCT FRAGILITY 1213

modiﬁcations have resulted in substantial cost savings in

packaging materials, reduced damage in shipment, and

potentially large savings from improved logistics and

shipping costs.

There are many potential hazards in the distribution

environment, including shock, vibration, temperature or

humidity extremes, electrostatic discharge, magnetic

ﬁelds, and compression. This article deals primarily with

shock and vibration inputs; however, it is important to

explore carefully those areas where the product is sensi-

tive or a large environmental input is likely.

The steps involved in designing an optimized package

system are as follows:

1. Deﬁne the environment in terms of shock and

vibration inputs likely during the product’s manu-

facturing and distribution cycle.

2. Deﬁne product fragility in terms of shock and

vibration.

3. Obtain product improvement feedback; that is,

examine product improvements in light of the eco-

nomic tradeoffs between packaging/distribution

costs and product modiﬁcation costs.

4. Evaluate cushion material performance; that

is, evaluate the shock and vibration characteristics

of available cusion materials and systems.

5. Design the package system; that is, select cushion

materials and design for optimum shock and vibra-

tion performance and minimum cube.

6. Test the ﬁnal package; that is, verify that the

package system performs as designed and properly

protects the product.

The primary purpose of this article is to examine

product fragility testing in detail, and therefore the other

steps are not covered in a comprehensive fashion. Refer to

the bibliography for excellent background material on the

other steps in this process (1–6) (see also Cushioning,

design; Distribution hazards).

DEFINING THE ENVIRONMENT

The end result of deﬁning the environment should be the

establishment of a design drop height and a resonant-

frequency spectrum for a particular product in a given

distribution environment. That is, based on the size and

weight of a packaged product, the engineer selects a drop

height that represents a given probability of input. Also,

vibration proﬁles for likely modes of transportation are

selected. Both of these pieces of information are used in

the last step, testing of the package system (7–10).

DEFINING PRODUCT FRAGILITY

The term product fragility is misunderstood by many

people and often conjures up images of totally destroyed

products, broken bottles, and the like. In reality, product

fragility is simply another product characteristic such as

size, weight, and color. Just as other product characteristics

are determined by measurement, product fragility (or

product ruggedness) can be measured with shock inputs.

This measurement takes the form of a damage-boundary

curve for shock and resonant frequency plots for vibration.

In both cases, the importance of determining these char-

acteristics cannot be overemphasized. Most people would

not think of buying a pair of shoes based on guessing their

foot size. It is just as short-sighted to design a package

system by guessing at product fragility.

Shock Fragility Assessment

The damage boundary is the principal tool used to deter-

mine the ruggedness of a product; it takes the general

shape shown in Figure 1. This plot 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 will likely

cause damage to the product regardless of whether it is

packaged or not. (This is a product test.)

Acceleration is a vector quantity describing the time

rate of change of velocity of a body in relation to a ﬁxed

reference point; that is, it describes the rate at which

velocity is increasing (acceleration) or decreasing (decel-

eration). The terms acceleration and deceleration are often

used interchangeably because most products respond

similarly to a rapid start or a rapid stop. Both terms are

expressed in G values, multiples of earth’s gravitational

constant, g.

The SI units

are m/s

Velocity change is the difference in a system’s velocity

magnitude and direction from the start to the end of shock

pulse. Velocity change is the integral of the acceleration

versus time pulse and is directly related to drop height.

To run a damage boundary, mount the product to be

tested on the table of a suitable shock-test machine (see

Figure 2). Secure the product with a rigid ﬁxture that

lends even support to the product over its entire surface.

Figure 1. Damage boundary (single orientation). A

, critical

acceleration; V

, critical velocity. To convert in. to cm, multiply

by 2.54.

1214 TESTING, PRODUCT FRAGILITY

The ﬁxture must be as rigid as possible so that it does not

distort the shock pulse transmitted to the product.

Set the shock machine to produce a low velocity-change

shock pulse with a duration of approximately 2 ms (a half-

sine waveform is generally used for this test). After the

shock pulse is delivered to the product, examine it to

determine if damage has occurred. If not, set the shock

machine to produce a slightly higher velocity change and

repeat the test. Continue this process with small incre-

mental increases in velocity change until damage occurs.

The last nonfailure shock input deﬁnes critical velocity

change DV

for the product in that orientation.

Next, fasten a new test specimen to the shock table and

set the machine to produce a trapezoidal pulse with a low

acceleration level and a velocity change at least two times

that of the critical velocity determined in the previous test.

Program the shock pulse into the product and, as before,

examine the product to determine if failure occurs. If no

failure has occurred, set the shock machine to produce a

higher acceleration level at approximately the same velo-

city change. Repeat this procedure with small increments

in acceleration until the failure level is reached. The last

nonfailure shock input deﬁnes the critical acceleration, A

for the product in that orientation (11).

The damage boundary curve may now be plotted by

drawing a vertical line through the critical velocity change

point and a horizontal line through the critical accelera-

tion point. The intersection of these two lines (the knee) is

a smooth curve as shown in Figure 1. A rectangular corner

may be used as a conservative approximation of the

damage region (12).

Note that the critical acceleration as determined by a

trapezoidal pulse is conservative when compared to that

generated through the use of other waveforms; that is, a

trapezoidal pulse is more damaging than other waveforms

of the same peak acceleration and duration. This is shown

graphically in Figure 3.

Since the shape of the waveform transmitted through

various packaging materials during impact is generally

not known, the use of a trapezoidal wave during damage

boundary testing results in a higher conﬁdence level in a

packaging system and is recommended for this purpose. It

should also be pointed out that the use of the trapezoidal

waveform results in a nearly linear abscissa on the

damage boundary. This means that it is necessary to

determine only one point on that axis in order to deter-

mine the critical acceleration for the product in that

orientation. Other waveforms result in critical accelera-

tions that are a complex function of the natural frequency

of components within the product. The procedure de-

scribed is based on ASTM D3332 (13).

The damage boundary is a valuable and powerful tool.

Critical velocity change is equated to equivalent free

drop height from the formula DV ¼ð1 þ eÞ

ﬃﬃﬃﬃﬃﬃﬃﬃﬃ

2gh

, where e

is the coefﬁcient of restitution of the impact surfaces (V

), g is acceleration of gravity [386 in./s

at sea level

(9.8 m/s

)], and h is equivalent free-fall drop height in in.

(or m). The critical velocity change tells the designer how

high the unpackaged product can fall onto a rigid surface

before damage occurs in that axis. If this equivalent drop

Figure 2. Shock-test machine.

Figure 3. Damage boundary for various pulses.

TESTING, PRODUCT FRAGILITY 1215

height is likely to be exceeded in the distribution environ-

ment, then the product must be cushioned. The perfor-

mance requirements of the cushion are that no more than

the critical acceleration be transmitted to the product.

In theory, the value of e (coefﬁcient of restitution) can

vary from zero to one. A value of zero would imply a totally

plastic impact with no rebound whatsoever. A value of one

implies a perfectly elastic impact where the rebounded

velocity is exactly equal to the impact velocity. As a

practical matter, a range of e = 0.25–0.75 produces good

accuracy in calculating the range of equivalent freefall

drop height. The chart in Figure 4 shows the effect of e.

The damage boundary also tells the engineer that at

low velocity changes, inﬁnite accelerations are possible

and that at low acceleration levels, inﬁnite velocity

changes are possible without product damage. That

means that it is necessary to deﬁne both critical accelera-

tion and critical velocity change to properly characterize

the fragility (or ruggedness) of a product.

Before running the damage-boundary test, the engi-

neer must deﬁne what constitutes damage to the product.

On one extreme, damage may be catastrophic failure.

However, there are many less severe damage modes that

can make a product unacceptable to the customer. In some

cases, damage can be determined by observation of the

product; at other times it involves running sophisticated

functional checks. Once the determination of damage is

made, the deﬁnition must remain constant throughout the

testing and must be consistent with what is deemed

unacceptable to the customer.

In general, damage-boundary tests must be run for

each axis in each orientation of the product. In the case of

a rectangular product such as a television set, this means

that a total of 12 specimens are necessary for a rigorous

test. However, since this testing should be done in the

product prototype stage, this quantity is rarely available

for a potentially destructive test. As a practical matter,

much information can be gained from a limited number of

samples, and multiple damage boundaries are often run

on the same unit in different orientations.

Vibration-Fragility Assessment

Determining product-vibration sensitivities involves iden-

tifying resonant frequencies of components within the

product in each of the principal axes. Resonance is that

characteristic displayed by all spring/mass systems

wherein at a given frequency, the response acceleration

of a component is greater than the input acceleration. This

characteristic is shown graphically in Figure 5.

As a general rule, the product will not be damaged due

to nonresonant inertial loading caused by vibration in the

distribution environment. This is because the acceleration

levels of most vehicles are relatively low when compared to

the critical acceleration of most products. It is only when a

component within a product is excited or forced by vibra-

tion at its natural frequency that damage is likely to occur.

At frequencies below the resonant frequency, the re-

sponse of a critical component is roughly equal to the

input (the response/input ratio is 1). At frequencies higher

than the resonant frequency, the response acceleration is

lower than the input. In this region a component acts as

its own isolator and results in a condition known as

attenuation.

At (and near) the product resonant frequency, however,

the response acceleration can be very much greater than

the input, causing product fatigue and ultimate failure in

a relatively short time. The purpose of vibration sensitiv-

ity assessment is to identify those critical frequencies

likely to cause damage to the product.

A resonant frequency search test is run by attaching a

product to the table of a suitable vibration test machine

(Figure 6) and subjecting it to an acceleration input at a

low level (typically 0.25–0.5 G) over a suitable frequency

Figure 4. Coefﬁcients of restitution. To convert inches to centi-

meters, multiply by 2.54.

Figure 5. Resonant frequency plot.

1216 TESTING, PRODUCT FRAGILITY

range, typically 1–300 Hz (cycles per second). Acceler-

ometers are fastened to critical components within the

product in order to determine the components’ responses

to the input acceleration. The response/input ratio is

plotted as a function of frequency for each ‘‘critical’’

componant. This ratio reaches a maximum at the compo-

nent resonant frequency. The test usually involves mon-

itoring many components in each axis of the product in

order to characterize its overall vibration sensitivities

(14). Sinusoidal or random excitation can be utilized for

this test, but random vibration excitation has gained wide

popularity due to the ease of use, test efﬁciency, and

potentially more accurate data.

The importance of vibration teting cannot be over-

emphasized. Any product shipped from point A to point

B is subject to vibration because of the transit vehicle it is

riding. The probability of this input is 100%. In contrast,

the probability of a shock input because of a drop is exactly

that, a probability function. In some cases, the drop height

experienced by a product may be severe, in most cases, it is

hardly measurable. However, any product that is shipped

in a vehicle is subject to vibrational input and it should be

tested for sensitivity to that input.

CONCLUSION

At this point, the engineer has sufﬁcient data to make

intelligent decisions about tradeoffs between the product

modiﬁcations and package costs. If a fragile component

can be ruggedized at minimal cost, resulting in

substantial package savings, then it makes sense to

pursue the product modiﬁcation.

The package-design process uses environmental data,

product-fragility information, cushion-performance data

(see Foam cushioning; Testing, cushion system), and a

healthy dose of designer creativity. Knowledge of package-

fabrication techniques, as well as other vital information

on ﬂammability restrictions, maximum weight and cube

for storage and transportation, recyclability of the pack-

age components, future cost trends of various key materi-

als, and so on. (1, 3, 5, 6, 15), is essential.

Once the design has been ﬁnalized and a prototype

fabricated, it must be tested to verify compliance with

product requirements. It is important to specify the cor-

rect inputs, both their magnitude (or duration) and se-

quence, in order to closely duplicate the potentially

damaging effects of the distribution environment. Test

procedures such as ASTM D4169 (7) and a variety of new

ISTA test procedures (17) have done much for improving

the correlation between laboratory tests and ﬁeld experi-

ence (2, 4, 16).

A properly engineered packaged system is now within

reach of all manufacturers, distributors, and other pack-

age-system users. The wasteful practice of overpackaging

can virtually disappear along with damage-in-shipment

reports. The tools are available and the technology is

straightforward. Optimized packaging is indeed an attain-

able goal.

BIBLIOGRAPHY

1. M. E. Gigliotti, Design Criteria for Plastic Package Cushion-

ing Materials, Plastic Technical Evaluation Center, Picatinny

Arsenal, Dover, NJ, 1962.

2. T. J. Grabowski, Design and Evaluation of Packages Contain-

ing Cushioned Items Using Peak Acceleration Versus Static

Stress Data, Shock, Vibration and Associated Environments

Bulletin, No. 39, part II, Ofﬁce of the Secretary of Defense,

Research and Engineering, Washington, DC, 1962.

3. C. Henny and F. Leslie, An Approach to the Solution of Shock

and Vibration Isolation Problems as Applied to Package

Cushioning Materials, Shock, Vibration and Associated En-

vironments Bulletin, No. 30, part II, Ofﬁce of the Secretary of

Defense, Research and Engineering, Washington, DC, 1962.

4. M. T. Kerr, ‘‘The Importance of Package Testing in Today’s

Data Processing Industry’’ in Proceedings of Western Regional

Forum Packaging Institute, USA, 1981.

5. R. D. Mindlin, Bell System Tech. J. 24(304), 352 (1945).

6. S. Mustin, Theory and Practice of Cushion Design, Shock and

Vibration Information Center, U.S. Department of Defense

Washington, DC 1968.

7. ASTM D4169-05, Standard Practice for Performance Testing

of Shipping Containers and Systems, American Society for

Testing and Materials, Philadelphia, 2005.

8. S. G. Guins, in Shock and Vibration Handbook, Part II,

McGraw-Hill, New York, 1961, Chapter 45.

9. F. E. Ostrem and B. Libovicz, A Survey of Environmental

Conditions Incident to the Transportation of Materials, Report

PB-204 442, General American Transportation Corp., Niles,

IL, 1971.

Figure 6. Vibration-test machine.

TESTING, PRODUCT FRAGILITY 1217