ASM Metals HandBook Vol. 8 - Mechanical Testing and Evaluation

Подождите немного. Документ загружается.

Fig. 18 Graphic representation of creep data showing various ways to plot time-dependent strain in

response to time-dependent stress. See text for discussion.

Creep Modulus. In addition to stress-strain plots versus time, creep behavior is also expressed as a creep

modulus, E(t), where

E(t) = σ/ε(t)

where σ is the applied stress and ε(t) is the creep strain as a function of time. The creep modulus is a measure of

rigidity that can be applied for tensile, shear, compressive, or flexural load conditions. However, the creep

modulus E(t) is neither a design property nor a material constant. It is a time-dependent variable that is also a

function of temperature and environment. The use of creep modulus data requires definition of intended design

life and test conditions that accurately reflect the intended application.

Creep Rupture. Like the creep modulus, creep rupture data depend strongly on temperature. Creep rupture, in

many respects, is a more important parameter because it represents the ultimate lifetime of a given material.

Two types of graphic representation can be constructed for the creep rupture envelope, as shown in Fig. 18(b)

and 18(c). Figure 18(b) shows a semilog plot of creep rupture stress as a function of failure time. For most

plastic candidates for long-term performance, the design life can be quite long—months or years. As a result,

the log-log coordinate system (Fig. 18c) has greater utility. Furthermore, creep rupture data tend to display

linearly on this coordinate scheme.

Creep strain data plots can be done in various forms. Figure 18 shows three methods of analyzing these data.

Each method holds one variable (stress, strain, or time) to be constant. For constant stress, Fig. 18(d) and 18(e)

apply. The data can be displayed either as a set of (usually near-linear) linear lines on log-log paper (Fig. 18d)

or as curvilinear lines on semilog paper (Fig. 18e). The parallel straight lines on log-log coordinates are called a

creep strain plot (Fig. 18d).

Isochronous Creep Data. If the time parameter is held constant, a set of isochronous (or constant time) stress-

strain curves results (Fig. 18f). A linear coordinate system is used to display these results. The slopes of these

isochronous creep curves produce the isochronous modulus graph (Fig. 18g). If the slopes of the semilog curves

are replotted against time, a set of nearly linear lines on semilog paper results. This represents the time-

dependent creep modulus plot (Fig. 18h). Most creep design data published in the United States are reported in

this manner.

Isometric Creep Data. If the strain is constant, isometric creep curves (Fig. 18i) result. The graph is usually

semilog in time. Isometric creep data are used extensively in Europe. The isometric modulus data can also be

extracted from these curves.

Dynamic Mechanical Properties

Dynamic mechanical tests measure the response of a material to a sinusoidal or other periodic stress. Because

of the viscoelastic nature of plastics, the stress and strain are generally not in phase. Two quantities, a stress-to-

strain ratio and a phase angle, are measured. Because dynamic properties are measured at the small deformation

around the equilibrium position, they involve only relative displacement of polymer chains in the linear-

response region. This measurement offers considerable information on structural property relationships, ranging

from local interaction of segments to the macrostructure of polymer chains.

Dynamic mechanical tests give a wider range of information about a material than other short-term tests

provide, because test parameters, such as temperature and frequency, can be varied over a wide range in a short

time. Superposition of data from different temperatures is also possible. Dynamic data can be interpreted from

the chemical structure and physical aggregation of the material. The results of dynamic measurements are

generally expressed as complex modulus, which is defined as the following:

G* = G′ + iG″

or through the dissipation factor, tan δ, which is related to complex moduli by:

Molecular weight, cross linking, crystallinity, and plasticization can affect the dynamic modulus. As a general

rule, these factors affect G′ the same way they affect complex modulus. In fact, one can convert shear modulus

to complex modulus, and vice versa, at least from a theoretical point of view.

The dissipation loss factor is generally an indication of reduced dimensional stability. A material with a high

loss factor, however, is useful for acoustical insulation. The dissipation factor shows a peak when there is a

phase transition. Thus, it is a sensitive method for detecting the existence of transitions. Low-temperature

transitions measured by this technique are related to high-impact properties for materials such as polycarbonate.

Impact Toughness

As would be expected, the impact toughness of plastics is affected by temperature. At temperatures below the

glass transition temperature, T

, the material is brittle, and impact strength is low. The brittleness temperature

decreases with increasing molecular weight. This is the reverse of the effect of molecular weight on the T

When the temperature increases to near the T

, the impact strength increases. With notch-sensitive materials

such as some crystalline plastics, environmental factors can create surface microcracks and reduce impact

strength considerably. Table 9 is a summary of fracture behavior of various plastics.

Table 9 Fracture behavior of selected plastic as a function of temperature

Temperature, °C ( °F)

Plastics

-20 (-4)

-10 (14)

0 (32)

10 (50)

20 (68)

30 (85)

40 (105)

50 (120)

Polystyrene A A A A A A A

Polymethyl methacrylate A A A A A A A

Glass-filled nylon (dry) A A A A A A A

Polypropylene A A A A B B B

Polyethylene terephthalate B B B B B B B

Acetal B B B B B B B

Nylon (dry) B B B B B B B

Polysulfone B B B B B B B

High-density polyethylene B B B B B B B

Rigid polyvinyl chloride B B B B B B C

Polyphenylene oxide B B B B B B C

Acrylonitrile-butadiene-styrene B B B B B B C

Polycarbonate B B B B C C C

Nylon (wet) B B B C C C C

Polytetrafluoroethylene B C C C C C C

Low-density polyethylene C C C C C C C C

(a) A, brittle even when unnotched; B, brittle, in the presence of a notch; C, tough

Although a number of standard impact tests are used to survey the performance of plastics exposed to different

environmental and loading conditions, none of these tests provides real, geometry-independent material data

that can be applied in design. Instead, they are only useful in application to quality control and initial material

comparisons. Even in this latter role, different tests will often rank materials in a different order. As a result,

proper test choice and interpretation require that the engineer have a very clear understanding of the test and its

relationship to specific design requirements.

Because of differing engineering requirements, a wide variety of impact test methods have been developed.

There is no one ideal method. The general classes of impact tests are shown in Fig. 19. However, this section

briefly describes three of the most commonly used tests for impact performance: the Izod notched-beam test,

the Charpy notched-beam test, and the dart penetration test.

Fig. 19 Categories of impact test methods used in testing of plastics. Source: Ref 4

Charpy Impact Test (ASTM D 256 and ISO 179). The Charpy geometry consists of a simply supported beam

with a centrally applied load on the reverse side of the beam from the notch (Fig. 20a). The notch serves to

create a stress concentration and to produce a constrained multiaxial state of tension a small distance below the

bottom of the notch. The load is applied dynamically by a free-falling pendulum of known initial potential

energy. The important dimensions of interest for these tests include the notch angle, the notch depth, the notch

tip radius, the depth of the beam, and the width of the beam. All these quantities, as well as more detailed

information specifying loading geometry and conditions, are described in ASTM D 256 and ISO 179.

Fig. 20 Specimen types and test configurations for pendulum impact toughness tests. (a) Charpy method.

(b) Izod method

Izod Impact Test (ASTM D 256 and ISO 180). Like the Charpy test, the Izod test involves a pendulum impact,

but the Izod geometry consists of a cantilever beam with the notch located on the same side as the impact point

(Fig. 20b). Because the pendulum hits the unnotched side of the sample in the Charpy test, Charpy values may

be much higher impact strength values than Izod test values. However, the two measurements can be correlated

(Ref 14). The Izod test is usually done on 3.2 mm (⅛ in.) thick samples. Materials such as polycarbonate

exhibit thickness-dependent impact properties. Below 6.4 mm (¼ in.), this material is ductile with a very high

value, but above this thickness, the material has a much lower value.

Unnotched impact toughness tests in ASTM D 4812 and D 3029 (dart penetration test) have been replaced by

ASTM D 5420. Impact values with unnotched samples are often considered a more definitive measure of

impact strength, while the Izod test indicates notch sensitivity.

Dart Penetration (Puncture) Test. Another impact test that is often reported is the dart penetration (puncture)

test. This test (Fig. 21) is different from the Izod and Charpy tests in a number of aspects. First, the stress state

is two-dimensional in nature because the specimen is a plate rather than a beam. Second, the thin, platelike

specimen does not contain any notches or other stress concentrations.

Fig. 21 Schematic puncture test geometry

The geometry and test conditions often applied using this specimen were described in ASTM D 3029 (now

replaced by ASTM D 5420). The quantity most often quoted with respect to this test is the energy required for

failure. Of course, these energy levels are very different from the notched beam energies-to-failure, but they

also do not represent any fundamental material property. A marked transition in mode of failure can also be

observed with this specimen as the rate is increased or the temperature is decreased. However, this transition

temperature is quite different from that measured in the notched beam tests. Usually it displays a transition form

ductile to brittle behavior at much lower temperatures than the notched specimens.

The dart penetration test is often performed with different specimens and indenter geometries. Linear elastic,

small-displacement, thin-plate theory has occasionally been used to analyze test results in an effort to compare

the performance of different materials tested with different specimen geometries. In all but the most brittle

materials, this is an inappropriate simplification of the test. A number of very nonlinear events can take place

during this test, including a growing indenter contact area, yielding, and large-displacement and large-strain

deformation. References 15, 16, 17 provide more details on these events and their effects on the test data.

Fracture Mechanics. Another way to evaluate the toughness of materials is by fracture toughness testing, where

the value of the critical stress-intensity factor for a material can be measured by testing standard cracked

specimens, such as the compact-tension specimen. Standard test methods and specimen geometries are defined

for measuring the critical stress-intensity factors for metals (ASTM E 399), but similar standards have yet to be

officially defined for plastics. It appears that many of the recommendations of the ASTM E 399 test procedure

for metals are equally worthwhile for plastics, although the ductile nature and low yield strength of plastics

pose problems of specimen size.

In fracture toughness testing, the sample size can be reduced as long as all dimensions of the laboratory

specimen are much larger than the plastic zone size. In fracture testing, according to ASTM E 399, the

thickness, B, must be:

where K

is the plane-strain fracture toughness, σ

is the yield stress, and r

is the radius of the plastic zone,

which is given by:

By ensuring that the thickness is much larger than the yield zone size (at least 16 times larger), the laboratory

specimen will be in the state of plane strain. Because of the hydrostatic stresses that develop at crack tips under

plane-strain conditions, yielding is suppressed, and a minimum value for fracture toughness is obtained. The

plane-strain fracture toughness can be used with confidence in designing large components.

Similar arguments hold for polymer fracture testing. To design large polymer components or to design for

polymer applications in which yielding is suppressed, it is important to measure fracture properties under

conditions of plane strain. However, typical engineering plastics have fracture toughnesses in the range of 2 to

4 MPa (1.8–3.6 ksi ) and yield strengths in the range of 50 to 80 MPa (7.3–11.6 ksi) (Ref 18). Plane-

strain testing conditions would require sample thicknesses in the range of 1.6 to 16 mm (0.06–0.63 in.). The

low-end range is a common size range, but the high end is more questionable. Engineering components

designed with polymers almost never use polymers as thick as 16 mm (0.63 in.). Therefore, it is not clear that

the plane-strain fracture toughness is the appropriate design data for engineering components in which the

polymers will experience only plane-stress conditions. More importantly, fabricating thick polymer samples for

plane-strain testing presents significant difficulties. Engineering plastics with fracture toughnesses in the range

of 2 to 4 MPa (1.8–3.6 ksi ) are not particularly tough. Rubber-toughened polymers can have much

higher toughness. Also, because the yield strength of rubber-toughened polymers is usually lower, the thickness

requirements for plane-strain fracture testing are such that potential laboratory specimens cannot be prepared.

Some of these toughened polymers can be tested with J integral techniques adapted from the J integral metals

standard (Ref 19, 20). Another technique known as the essential work of fracture technique has been

considered. It has the potential to provide both plane-stress and plane-strain fracture toughness results for

polymers. The essential work of fracture data can be obtained on thin polymers having thicknesses similar to

those of typical polymer components (Ref 21).

Hardness Tests

Typical hardness values of common plastics are listed in Table 10. Rockwell testing and the durometer test

method are the most common, although another type of hardness test for plastics is the Barcol method. A rough

comparison of hardness scales for these methods is in Fig. 22, but it must be understood that any conversions

from Fig. 22 are only rough estimates that vary depending on the materials. Hardness conversions are

complicated by several material factors such as elastic recovery and, for plastics, the time-dependent effects

from creep behavior. More information on the hardness testing of plastics is also given in the article “Selection

and Industrial Applications of Hardness Tests” in this Volume.

Table 10 Typical hardness values of selected plastics

Rockwell Plastic material

HRM HRR

Durometer,

Shore D

Barcol

Thermoplastic

Acylonitrile-butadiene-styrene … 75–115 …

…

Acetal 94 120 …

…

Acrylic 85–105 … …

…

Cellulosics … 30–125 …

…

Polyphenylene oxide 80 120 …

…

Nylon … 108–120 …

…

Polycarbonate 72 118 …

…

High-density polyethylene … … 60–70

…

Low-density polyethylene … … 40–50

…

Polypropylene … … 75–85

…

Polystyrene 68–70 … …

…

Polyvinyl chloride (PVC)(rigid) … 115 …

…

Polysulfone 70 120 …

…

Thermosets

Phenolic (with cellulose) 100–110 … …

…

Phenolic (mineral filler) 105–115 … …

…

Unsaturated polyester (clear cast) … … …

34–40

Polyurethane (high-density integral skin foam) … … 36–63

…

Polyurethane (solid reaction injection molded elastomer) … … 39–83

…

Epoxy (fiberglass rein forced) 106–108 … … …

Fig. 22 Approximate relations among hardness scale for plastics

Rockwell hardness tests of plastics (ASTM D 785 and ISO 2039) are ball-indentation methods, where hardness

is related to the net increase in the depth of an indentation after application of a minor load and a major load.

The ball diameter and the loads are specified for each of the Rockwell scales, which are R, L, M, E, and K in

order of increasing hardness. The Rockwell test is used for relatively hard plastics such as thermosets and

structural thermoplastics such as nylons, polystyrene, acetals, and acrylics. Typical Rockwell values are shown

in Fig. 23.

Fig. 23 Rockwell hardness of engineering plastics. PET, polyethylene terephthalate; PA, polyamide;

PPO, polyphenylene oxide; PBT, polybutylene terephthalate; PC, polycarbonate; ABS, acrylonitrile-

butadienestyrene

The durometer (or Shore hardness) method (ASTM D 2240 and ISO 868) registers the amount of indentation

caused by a spring-loaded pointed indenter. This method is used for softer plastics and rubbers, and 100 is the

highest hardness rating of this scale. Two types of durometers are used: type A and type D, as described further

in the article “Selection and Industrial Applications of Hardness Tests” in this Volume.

The Barcol hardness test (ASTM D 2583) is mainly used for measuring the hardness of reinforced and

unreinforced rigid plastics. A hardness value is obtained by measuring the resistance to penetration of a sharp

steel point under a spring load. The instrument, called the Barcol impressor, gives a direct reading on a 0 to 100

scale. The hardness value is often used as a measure of the degree of cure of a plastic.

International Rubber Hardness Degrees (IRHD) Testing. The IRHD hardness test is very similar to durometer

testing with some important differences. Durometer testers apply a load to the sample using a calibrated spring

and a pointed or blunt shaped indenter. The load therefore will vary according to the depth of the indentation,

because of the spring gradient. The IRHD tester uses a minor-major load system of constant load and a ball

indenter to determine the hardness of the sample. This method is described further in the article “Miscellaneous

Hardness Tests” in this Volume.

Fatigue Testing

Compared to testing of metals, the testing of plastics is a relatively recent pursuit. Because engineers and

designers always use knowledge gained from previous experience, the methods used to test plastics in fatigue

are largely based on methods developed for metals, with accommodations to account for the more obvious

differences between the two materials.

For example, as previously noted in the section “Dynamic Mechanical Properties,” the role of high hysteresis

losses in the repeated stressing of plastics is very important. Unlike metals, plastics deform in a largely

nonelastic manner, resulting in part of the mechanical energy being converted into heat within the material. The

gradual buildup of heat may be sufficient to cause a loss in strength and rigidity. This effect is further

aggravated by the low thermal conductivity of plastics and a general increase in hysteresis losses with an

increase in temperature. Hysteresis losses are also a function of the loading rate (frequency), the type of load

(bending, tension, or torsion), and the volume of material under stress. The hysteresis losses increase with

loading rate and the volume of material under stress.

This also can be further extended to include the effects of different loading waveforms (sinusoid, saw tooth, or

square) on the fatigue strength of viscoelastic materials. In addition, absorbed water and environmental

variables also influence the fatigue strength of plastics. These and other factors are described in more detail in

the article “Fatigue Testing and Behavior of Plastics” in this Volume.

References cited in this section

4. F.N. Kelly and F. Bueche, J. Polym. Sci., Vol 50, 1961, p 549

5. V. Shah, Handbook of Plastics Testing Technology, 2nd ed., John Wiley & Sons, 1998

• ISO/IEC Selected Standards for Testing Plastics, 2nd ed., ASTM, 1999

12. J. Nairn and R. Farris, Important Properties Divergences, Engineering Plastics, Vol 2, Engineered

Materials Handbook, ASM International, 1988, p 655–658

13. T. Osswald, Polymer Processing Fundamentals, Hanser/Gardner Publications Inc., 1998, p 19–43

14. K.M. Ralls, T.H. Courtney, and J. Wulff, Introduction to Materials Science and Engineering, John

Wiley & Sons, 1976

15. S. Turner, Mechanical Testing, Engineering Plastics, Vol 2, Engineered Materials Handbook, ASM

International, 1988, p 547

16. R.B. Seymour, Polymers for Engineering Applications, ASM International, 1987, p 155

17. V. Shah, Handbook of Plastics Testing Technology, 2nd ed., John Wiley & Sons, 1998

18. Recognized Components Directories, Underwriters Laboratories

19. R.D. Deanin, Polymer Structure, Properties and Applications, Cahners, 1982

26. R.P. Nimmer, Analysis of the Puncture of a Polycarbonate Disc, Polym. Eng. Sci., Vol 23, 1983, p 155

27. R.P. Nimmer, An Analytical Study of Tensile and Puncture Test Behavior as a Function of Large-Strain

Properties, Polym. Eng. Sci., Vol 27, 1987, p 263

28. L.M. Carapelucci, A.F. Yee, and R.P. Nimmer, Some Problems Associated with the Puncture Testing of

Plastics, J. Polym. Eng., June 1987

29. J.G. Williams, Fracture Mechanics of Polymers, Ellis Horwood, 1984

30. D.D. Huang and J.G. Williams, J. Mater. Sci., Vol 22, 1987, p 2503

31. M.K.V. Chan and J.G. Williams, Int. J. Fract., Vol 19, 1983, p 145

32. Y.W. Mai and B. Cotterell, Int. J. Fract., Vol 32, 1986, p 105

Mechanical Testing of Polymers and Ceramics

Elastomers and Fibers

As previously noted, polymers can exhibit a range of mechanical behaviors that characterize their various

classifications as elastomers, plastics, and fibers (Fig. 1). The following discussions briefly describe tension

testing of elastomers and fibers.

Tension Testing of Elastomers

Elastomers have the ability to undergo high levels of reversible elongation that, in some cases, can reach up to

1000%. This high degree of reversible elongation allows stretching and recovery like that of a rubber band.

Over 20 different types of polymers can be used as bases for elastomeric compounds, and each type can have a

significant number of contrasting subtypes within it. Properties of different polymers can be markedly different:

for instance, urethanes seldom have tensile strengths below 20.7 MPa (3.0 ksi), whereas silicones rarely exceed

8.3 MPa (1.2 ksi). Natural rubber is known for high elongation, 500 to 800%, whereas fluoroelastomers

typically have elongation values ranging from 100 to 250%.

Literally hundreds of compounding ingredients are also available, including major classes such as powders

(carbon black, clays, silicas), plasticizers (petroleum-base, vegetable, synthetic), and curatives (reactive

chemicals that change the gummy mixture into a firm, stable elastomer). A rubber formulation can contain from

four or five ingredients to 20 or more. The number, type, and level of ingredients can be used to change

dramatically the properties of the resulting compound, even if the polymer base remains exactly the same.

ASTM D 412 is the U.S. Standard for tension testing of elastomers. It specifies two principal varieties of

specimens: the more commonly used dumbbell-type, die cut from a standard test slab 150 mm by 150 mm by

20 mm (6 in. by 6 in. by 0.8 in.), and actual molded rings of rubber. The second type was standardized for use

by the O-ring industry. For both varieties, several possible sizes are permitted, although, again, more tests are

run on one of the dumbbell specimens (cut using the Die C shape) than on all other types combined. Straight

specimens are also permitted, but their use is discouraged because of a pronounced tendency to break at the grip

points, which makes the results less reliable.

The power-driven equipment used for testing is described, including details such as the jaws used to grip the

specimen, temperature-controlled test chambers when needed, and the crosshead speed of 500 mm/min (20

in./min). The testing machine must be capable of measuring the applied force within 2%, and a calibration

procedure is described. Various other details, such as die-cutting procedures and descriptions of fixtures, are

also provided.

The method for determining actual elongation can be visual, mechanical, or optical, but the method must be

accurate within 10% increments. In the original visual technique, the machine operator simply held a scale

behind or alongside the specimen as it was being stretched and noted the progressive change in the distance

between two lines marked on the center length of the dogbone shape. The degree of precision that could be

attained using a handheld ruler behind a piece of rubber being stretched at a rate of over 75 mm/s (3 in./s) was

always open to question, with 10% being an optimistic estimate.

More recent technology employs extensometers, which are comprised of pairs of very light grips that are

clamped onto the specimen and whose motion is then measured to determine actual material elongation. The

newest technology involves optical methods, in which highly contrasting marks on the specimen are tracked by

scanning devices, with the material elongation again being determined by the relative changes in the reference

marks.

Normal procedure calls for three specimens to be tested from each compound, with the median figure being

reported. Provision is also made for use of five specimens on some occasions, with the median again being

used.

Techniques for calculating the tensile stress, tensile strength, and elongation are described for the different

types of test specimens. The common practice of using the unstressed cross-sectional area for calculation of

tensile strength is used for elastomers, as it is for many other materials. It is interesting to note that if the actual

cross-sectional area at fracture is used to calculate true tensile strength of an elastomer, values that are higher

by orders of magnitude are obtained.

In recent years, attention has been given to estimating the precision and reproducibility of the data generated in

this type of testing. Interlaboratory test comparisons involving up to ten different facilities have been run, and

the later versions of ASTM D 412 contain the information gathered.

Variability of the data for any given compound is to some degree related to that particular formulation. When

testing was performed on three different compounds of very divergent types and property levels, the pooled

value for repeatability of tensile-strength determinations within labs was about 6%, whereas reproducibility

between labs was much less precise, at about 18%. Comparable figures for ultimate elongation were

approximately 9% (intralab) and 14% (interlab).

Similar comparisons of the 100% modulus (defined in “Modulus of the Compound” later in this section) have

shown even less precision, with intralab variation of almost 20% and interlab variation of over 31%. This runs

counter to the premise that modulus should be more narrowly distributed than tensile strength, because tensile

strength and ultimate elongation are failure properties, and as such are profoundly affected by details of

specimen preparation. Because the data do not support such a premise, some other factor must be at work.

Possibly that factor is the lack of precision with which the 100% strain point is observed, but, in any case, it is

important to determine the actual relationship between the precision levels of the different property

measurements.