Kutz M. Handbook of materials selection

Подождите немного. Документ загружается.

6 FATIGUE TESTING 537

R-ratio = 0.05

Cycles to Failure

Maximum Stress, ksi

100

120

140

Ti-6-4

7075 alum

filled points

in environment

R-ratio=0 data band

from handbook

Fig. 14 Conventional S–N fatigue test data presented for example titanium

and aluminum alloy materials.

the endurance limit of a metallic material; in theory the endurance limit is de-

ﬁned as the applied stress level that corresponds to inﬁnite life, in practical terms

inﬁnite life is approximated in the laboratory as a million cycles. The usual

approach

is to test 30 specimens using an iterative, staircase method to converge

on the stress level corresponding to the endurance limit. The nominal stress step

chosen is approximately 2% of the ultimate tensile strength of the metallic ma-

terial. Although this is a test-intensive methodology, the technique is able to

clearly deﬁne the endurance limit to the accuracy given by the stress step.

The specimens used for fatigue crack propagation testing are similar to that

described previously when discussing fracture. The most common geometries

are the compact tension, C(T), and the middle cracked tension, M(T), specimens.

However the pin-loaded C(T) is most suitable for positive R ratio (e.g., tension)

since the cracked geometry is unstable in compression. Fatigue crack growth

(FCG) testing is described in detail in one of the longest and most complex

ASTM test standards, ASTM E647. The standard has grown in length as the

understanding of the fatigue crack growth process has deepened. During FCG

testing the goal is to determine crack growth rate as a function of stress intensity

factor range,

⌬K. A typical series of fatigue crack growth curves are shown in

Fig. 15 for an aluminum alloy subject to a number of different stress ratios. The

central regions of the crack growth curves can be observed to be relatively linear

as expected in the moderate growth rate Paris regime.

One of the most critical aspects of fatigue crack growth testing is measure-

ment of the crack length. The most common method is to use a traveling tele-

microscope mounted on a vernier scale for accurate measurement. Recent

advances in both crack length measurement and test automation

9,10

employ elas-

538 TESTING OF METALLIC MATERIALS

Fatigue crack growth rate da/dN, in/cycle

110

⌬K, ksi公in

7075-T6x (extrusion)

R = -0.5 (2 tests)

R = -0.25 (2 tests)

R = 0.1 (2 tests)

R = 0.8 (1 test)

-8

-7

-6

-5

-4

-3

Fig. 15 Series of fatigue crack growth curves for aluminum at various R ratios.

tic compliance and electric potential drop methods to measure crack length.

However, it is critical to note that these indirect, nonabsolute measurement tech-

niques both require periodic visual crack length measurement for each test to

ensure proper calibration. The quality of a laboratory’s nonvisual crack length

measurement technique and equipment is critical and often directly related to

the quality of the FCG data generated.

The conventional method for performing crack growth testing is to record

growth rate data with a ﬁxed loading cycle as the crack grows. More efﬁcient,

newer methods are now available that critically depend upon nonvisual crack

length measurement methods to control the applied loading. For instance, K

gradient techniques can be used where the quantity C

⫽ dK/K da is ﬁxed at

different levels for different portions of the test. This increases efﬁciency since

(a) it allows the generation of multiple crack growth data sets from the same

specimen and (b) data can be generated in short growth intervals resulting in

much faster data generation than with constant amplitude loading. However,

these advanced fatigue crack growth test techniques must be carefully applied

by experienced testing professionals who fully understand the limitations of this

accelerated approach and have the instrumentation available to accurately mea-

sure crack length nonvisually.

Most fatigue crack growth rate testing is performed at a ﬁxed load ratio since

this variable does not change for many structures. However, another method

6 FATIGUE TESTING 539

mild steel

(medium strength)

∆

K = 20 ksi公in, R=0.6

Fatigue crack length, inch

Fatigue crack growth rate, x10

–6

inch/cycle

1.5

2.0

2.5

3.0

3.5

RT, 40-70% RH

–60

F, dry air

RT, 10-20% RH

140 F, 10% RH

RT, 45% RH

RT, 95% RH

140 F, 85% RH

low humidities

-60

140

140 F

humidity

Fig. 16 Inﬂuence of environment on fatigue crack growth rate behavior using

a constant ⌬K testing approach.

gaining wide popularity is generating FCG data from the highest rates down to

near threshold (e.g., crack growth rate at a very low magnitude, a quantity es-

pecially useful for design purposes) using a constant K

max

, increasing K

min

load-

ing condition. Consequently when the test starts, the R ratio is low (0.1 or so)

and as the test progresses R increases until approximately 0.9 or more in the

near-threshold regime. Whereas conventional fatigue crack growth threshold test-

ing may take weeks or months to complete, this method results in a highly

accelerated test and can be performed in a matter of days. Again, care must be

taken to ensure that the choice of the variable test control parameters does not

affect the magnitudes or trends of the fatigue crack growth rate data obtained.

For design purposes, it is sometimes necessary to understand the potential

impact of external variables on FCG behavior. For this case, constant

⌬K testing

can be employed to determine effects difﬁcult and time consuming to detect if

a complete crack growth curve were generated. As observed in the data in Fig.

16, several regions of nominally constant

⌬K are sequentially applied with the

external environmental variable perturbed for each segment. This allows a better

assessment of the effect of the variable since a statistically relevant number of

crack growth rates are available for each segment. Although this is a highly

540 TESTING OF METALLIC MATERIALS

efﬁcient method to determine effects, the disadvantage is that the measurement

strictly applies to only one constant

⌬K level. Nevertheless, a similar approach

can be used to indirectly determine, for instance, the magnitude of internal re-

sidual stresses in a component by examining small changes in fatigue crack

growth rate.

The challenge in applying fatigue crack growth rate data is accounting for

the variable amplitude loading environment typically applied to a component in

service. Most of the analytical tools available are empirically based and require

some testing under the variable amplitude, spectrum load condition. In these

cases, it is not uncommon to perform a crack growth test using a repeated

variable amplitude spectrum derived from either service recorded or design load-

ing. The test specimens in these cases are typically more complex, reﬂecting

some design detail such as a fastener hole or weld. Rather than generating crack

growth rate data with these tests, the goal is typically to provide crack length

as a function of spectrum passes (or cycle count). Although this type of test is

widely performed in the aerospace and ground vehicle industries,

the myriad

of approaches used has stiﬂed standardization, and it is only now underway

within ASTM to develop a standard practice applicable to this highly complex

test.

7 OTHER MECHANICAL TESTING

There are a number of test methods and procedures that are related to the areas

described earlier but, in the spirit of brevity, have been omitted. Nevertheless,

several are sufﬁciently popular and warrant some mention, albeit brief.

Some highly specialized assessments other than tensile tests exist for deter-

mining the strength and ductility of a metallic material. One common application

for the bending strength test is the use of ﬂat metallic materials for spring

applications (ASTM E855). The two primary properties that this test provides

are the modulus of elasticity in bending and the bending yield strength (pro-

portional limit). Moreover bend tests used to assess ductility (ASTM E190 and

E290) do not provide quantitative results, but rather provide a deformed sample

suitable for close visual inspection to determine basic metallic material behavior.

The region of highest strain is typically inspected for possible cracking, rumples,

or orange peeling as a quality control assessment suitable for the shop ﬂoor.

High strain rate testing is another area where signiﬁcant effort has been ex-

pended to generate the constituent properties applicable in the high rate regime.

For instance, the split Hopkinson pressure bar deforms a sample at a high strain

rate (

⬎100%) while applying uniaxial stress to the sample. Similar methods have

been developed that also allow high rate torsional loading. These methods have

been successfully applied to a wide range of metallic materials.

Finally, numerous test techniques are also available for testing adhesion, fric-

tion, and wear of surfaces. These are critical issues from a design viewpoint

with new standardized methods being developed fairly regularly.

8 ENVIRONMENTAL CONSIDERATIONS

It is often necessary to perform mechanical testing under environmental condi-

tions other than ambient air. For instance, fatigue crack growth testing is often

performed in either dry, high humidity, or aqueous conditions. For aqueous con-

8 ENVIRONMENTAL CONSIDERATIONS 541

Fig. 17 Chamber around a ﬂat panel M(T) specimen for suspending jet fuel or other aqueous

environments while measuring fatigue crack growth rate behavior.

Fig. 18 Schematic of a compact tension specimen sealed with a plastic bag.

ditions, a setup as shown in Fig. 17 can be constructed to perform crack growth

rate measurements. Sealing against a ﬂat surface is relatively straightforward

provided the materials used in the chamber are impervious to the environment

contained.

Other strategies such as that shown in Fig. 18 can be used to seal the loading

pins in a compact tension specimen. This technique has proven very suitable in

the past for containing aqueous environments. Clear Teﬂon (PTFE) bags are

commercially available that allow the setup to also be subjected to temperatures

in the range of

⫺400 to 300⬚F. Simpler methods using supermarket polypropyl-

ene bags can also be used to contain conditioned air by introducing either des-

iccant or a hydrated sponge for the dry and humid conditions, respectively.

Although the engineer’s inclination might be to build a more elaborate environ-

mental chamber, a simple plastic bag with duct tape seals works sufﬁciently well

542 TESTING OF METALLIC MATERIALS

Fig. 19 Several examples of enclosures for both high- and low-temperature materials testing.

(Photos courtesy of the Instron Corporation and MTS Corporation)

to contain dry [⬍5% relative humidity (RH)] and humid (⬎95% RH) environ-

ments.

Finally, it is sometimes not simple to superimpose two rate-dependent pro-

cesses without special attention. For instance, care must be taken to ensure that

the time-dependent effects that often occur in an environment are properly

applied in the materials test. As an example, the deleterious effect of seawater

is gradually lost during a fatigue crack growth test as the frequency exceeds 1

Hz.

If a geometry is somewhat unusual and a setup does not ﬁt into existing

furnaces, it is sometimes necessary to construct a custom furnace from insulating

ﬁberboard so long as the temperature required is not too high. In this case, a

circulating fan is included along with a heater element and inexpensive temper-

ature controller. However, an inexpensive approach such as the one just de-

scribed will not perform well given an aggressive environment, higher pressure

applications or more highly elevated temperatures. For these cases, more expen-

sive and custom containments are typically required as shown in Fig. 19.

Thermomechanical fatigue (TMF) differs from the fatigue testing described

earlier since the temperature of the specimen is programmed to vary in a precise

manner relative to the mechanical loading. The challenge with this type of testing

includes (a) achieving and (b) controlling the thermal environment required. The

limiting factor is not usually with heating as induction furnaces, for instance,

can heat a sample very rapidly in a uniform manner. The cycling rate is typically

controlled by cooling and concerns regarding large thermal gradients exist if

cooling occurs too quickly. Recent work within ASTM has generated a draft

standard for TMF testing that is currently undergoing balloting.

Acknowledgments

The author would like to graciously acknowledge the wisdom and lessons gained

from working closely with two highly experienced colleagues and mentors at

Southwest Research Institute. Jack FitzGerald taught me that often the simplest

approach, the one that seemed too elementary and often at odds with engineering

education, was always the best approach when faced with something unusual or

REFERENCES 543

unknown in the laboratory. Furthermore, Andy Nagy exhibited incredible crea-

tivity by taking half-baked, malformed ideas and transforming them into real

devices that functioned ﬂawlessly in the test lab. Both of these consummate

professionals, now deceased, provided excellent role models and mentors for the

development of an inexperienced test engineer. Finally, gratitude is extended to

FTA Inc, Instron, and MTS for supplying some of the photographs used herein.

REFERENCES

1. H. Kuhn and D. Medlin (Eds.), ASM Handbook on Mechanical Testing and Evaluation, Vol. 8,

ASM International, Materials Park, OH, 2000.

2. 2000 Annual Book of ASTM Standards, Section Three: Metals Test Methods and Analytical

Procedures, Vol. 3.01, American Society for Testing and Materials, West Conshohocken, PA,

2000.

3. International Directory of Testing Laboratories: 2001 Edition, American Society for Testing and

Materials, West Conshohocken, PA, 2001.

4. G. F. Lucas and D. A. Stubbs (Eds.), Nontraditional Methods of Sensing Stress, Strain and

Damage in Materials and Structures, ASTM STP 1318, American Society for Testing and Ma-

terials, West Conshohocken, PA, 1997.

5. G. F. Lucas, P. C. McKeighan, and J. S. Ransom (Eds.), Nontraditional Methods of Sensing

Stress, Strain and Damage in Materials and Structures: Second Volume, ASTM STP 1323,

American Society for Testing and Materials, West Conshohocken, PA, 2001.

6. Speciﬁcation for Casing and Tubing, API Speciﬁcation 5CT, American Petroleum Institute, 6th

ed., Washington, D.C., October 1998.

7. J. A. Joyce, Manual on Elastic-Plastic Fracture Laboratory Test Procedures, ASTM Manual

Series MNL 27, American Society for Testing and Materials, West Conshohocken, PA, 1996.

8. A Guide for Fatigue Testing and the Statistical Analysis of Fatigue Data, ASTM STP 91A, 2nd

ed., American Society for Testing and Materials, Philadelphia, PA, 1963.

9. A. A. Braun, P. C. McKeighan, M. A. Nicolson, and R. P. Lohr (Eds.), Applications of Automation

Technology to Fatigue and Fracture Testing and Analysis: Fourth Volume, ASTM STP 1411,

American Society for Testing and Materials, West Conshohocken, PA, 2002.

10. J. J. Ruschau and J. K. Donald (Eds.), Special Applications and Advanced Techniques for Crack

Size Determination, ASTM STP 1251, American Society for Testing and Materials, Philadelphia,

1995.

11. P. C. McKeighan and N. Ranganathan (Eds.), Fatigue Testing and Analysis Under Variable

Amplitude Loading Conditions, ASTM STP 14xx, American Society for Testing and Materials,

West Conshohocken, PA, 2003.

545

CHAPTER 20

PLASTICS TESTING

Vishu Shah

Consultek

Brea, California

1 MECHANICAL PROPERTIES 547

1.1 Tensile Tests (ASTM D638,

ISO 527-1) 547

1.2 Flexural Properties

(ASTM D790, ISO 178) 548

1.3 Creep Properties 551

1.4 Stress Relaxation 552

1.5 Impact Properties 555

1.6 Fatigue Resistance 564

1.7 Hardness Tests 567

2 THERMAL PROPERTIES 568

2.1 Tests for Elevated Temperature

Performance 568

2.2 Brittleness Temperature

(ASTM D746, ISO 974) 574

3 ELECTRICAL PROPERTIES 574

3.1 Dielectric Strength

(ASTM D149, IEC 243-1) 575

3.2 Dielectric Constant and

Dissipation Factor

(ASTM D150, IEC 250) 576

3.3 Electrical Resistance Tests 577

3.4 Arc Resistance (ASTM D495) 578

4 WEATHERING PROPERTIES 582

4.1 Accelerated Weathering Tests 584

4.2 Outdoor Weathering of Plastics

(ASTM D1435) 587

BIBLIOGRAPHY 589

Not too long ago, the concept of testing was merely an afterthought of the

procurement process. But now, with the advent of science and technology, the

concept of testing is an integral part of research and development, product de-

sign, and manufacturing. The question that is often asked is ‘‘why test?’’ The

answer is simple. Times have changed. The manner in which we do things today

is different. The emphasis is on automation, high production, and cost reduction.

There is a growing demand for intricately shaped, high-tolerance parts. Con-

sumer awareness, a subject totally ignored by the manufacturers once upon a

time, is now a major area of concern. Along with the requirements, our priorities

have also changed. While designing a machine or a product, the ﬁrst order of

priority in most cases is safety and health. Manufacturers and suppliers are now

required to meet a variety of standards and speciﬁcations. Obviously, relying

merely on past experience and quality of workmanship is simply not enough.

The following are some of the major reasons for testing:

1. To prove design concepts

2. To provide a basis for reliability

3. Safety

HandbookofMaterialsSelection.EditedbyMyerKutz

546 PLASTICS TESTING

4. Protection against product liability suits

5. Quality control

6. To meet standards and speciﬁcations

7. To verify the manufacturing process

8. To evaluate competitors’ products

9. To establish a history for new materials

In the last two decades, just about every manufacturer has turned to plastics to

achieve cost reduction, automation, and high yield. The lack of history along

with the explosive growth and diversity of polymeric materials has forced the

plastics industry into placing extra emphasis on testing and developing a wide

variety of testing procedures. Through the painstaking efforts of various stan-

dards organizations, material suppliers, and mainly the numerous committees of

the American Society for Testing and Materials (ASTM), over 10,000 different

test methods have been developed.

The need for developing standard test methods speciﬁcally designed for plas-

tic materials originated for two main reasons. Initially, the properties of plastic

materials were determined by duplicating the test methods developed for testing

metals and other similar materials. The Izod impact test, for example, was de-

rived from the manual for testing metals. Because of the drastically different

nature of plastic materials, the test methods often had to be modiﬁed. As a result,

a large number of nonstandard tests were written by various parties. As many

as 8–10 distinct and separate test methods were written to determine the same

property. Such practice created total chaos among the developers of the raw

materials, suppliers, design engineers, and ultimate end users. It became increas-

ingly difﬁcult to keep up with various test methods as well as to comprehend

the real meaning of reported test values. The standardization of test methods

acceptable to everyone solved the problem of communication between devel-

opers, designers, and end users, allowing them to speak a common language

when comparing the test data and results.

In spite of the standardization of various test methods, we still face the prob-

lem of comprehension and interpretation of the test data by an average person

in the plastics industry. This is due to the complex nature of the test procedures

and the number of tests and testing organizations. The key in overcoming this

problem is to develop a thorough understanding of what the various tests mean

and the signiﬁcance of the result to the application being considered. Unfortu-

nately, the plastics industry has placed more emphasis on how and not enough

on why, which obviously is more important from the standpoint of comprehen-

sion of the test results and understanding the true meaning of the values. The

lack of understanding of the real meaning of heat deﬂection temperature, which

is often interpreted as the temperature at which a plastic material will sustain

static or dynamic load for a long period, is one such classic example of misin-

terpretation. In the chapters to follow, we concentrate on the signiﬁcance, inter-

pretation, and limitations of physical property data and test procedures. Finally,

a word of caution: It is extremely important to understand that the majority of

physical property tests are subject to rather large errors. As a general rule, the

error of testing should be considered

Ⳳ5%. Some tests are more precise than

others. Such testing errors come about from three major areas: one from the

1 MECHANICAL PROPERTIES 547

basic test itself, the second from the operators conducting the tests, and the third

from the variations in the test specimens. While evaluating the test data and

making decisions based on test data, one must consider the error factor to make

certain that a valid difference in the test data exists.

1 MECHANICAL PROPERTIES

The mechanical properties, among all the properties of plastic materials, are

often the most important properties because virtually all service conditions and

the majority of end-use applications involve some degree of mechanical loading.

Nevertheless, these properties are the least understood by most design engineers.

The material selection for a variety of applications is quite often based on me-

chanical properties such as tensile strength, modulus, elongation, and impact

strength. These values are normally derived from the technical literature pro-

vided by material suppliers. More often than not, too much emphasis is placed

on comparing the published values of different types and grades of plastics and

not enough on determining the true meaning of the mechanical properties and

their relation to end-use requirements. In practical applications, plastics are sel-

dom, if ever, subjected to a single, steady deformation without the presence of

other adverse factors such as environment and temperature. Since the published

values of the mechanical properties of plastics are generated from tests con-

ducted in a laboratory under standard test conditions, the danger of selecting

and specifying a material from these values is obvious. A thorough understand-

ing of mechanical properties, tests employed to determine such properties, and

the effect of adverse conditions on mechanical properties over a long period is

extremely important.

1.1 Tensile Tests (ASTM D638, ISO 527-1)

Tensile elongation and tensile modulus measurements are among the most im-

portant indications of strength in a material and are the most widely speciﬁed

properties of plastic materials. Tensile test, in a broad sense, is a measurement

of the ability of a material to withstand forces that tend to pull it apart and to

determine to what extent the material stretches before breaking. Tensile modulus,

an indication of the relative stiffness of a material, can be determined from a

stress–strain diagram. Different types of plastic materials are often compared on

the basis of tensile strength, elongation, and tensile modulus data. Many plastics

are very sensitive to the rate of straining and environmental conditions. There-

fore, the data obtained by this method cannot be considered valid for applications

involving load-time scales or environments widely different from this method.

The tensile property data are more useful in preferential selection of a particular

type of plastic from a large group of plastic materials and such data are of limited

use in actual design of the product. This is because the test does not take into

account the time-dependent behavior of plastic materials.

The tensile testing machine of a constant-rate-of-crosshead movement is used.

It has a ﬁxed or essentially stationary member, carrying one grip and a movable

member carrying a second grip. Self-aligning grips employed for holding the

test specimen between the ﬁxed member and the movable member prevent align-

ment problems. A controlled velocity drive mechanism is used. Some of the

commercially available machines use a closed-loop servo-controlled drive mech-

anism to provide a high degree of speed accuracy. A load-indicating mechanism