ASM Metals HandBook Vol. 8 - Mechanical Testing and Evaluation

Подождите немного. Документ загружается.

where ε

and ε

are the axial and transverse strains, respectively (ε

< 0). A sample calculation is provided in

ASTM D 3518.

Stress State and Failure Modes. In addition to the shear stress τ

, there exist normal stresses, σ

and σ

, as

shown in Fig. 19. Because the stresses σ

and σ

are tensile, the apparent shear stress at failure may be reduced.

It has furthermore been shown that thin [±45°]

specimens with n ≤ 2 may fail at load and strain levels that are

too low and not representative for the shear strength of the material (Ref 32, 33). The thickness effect is

attributed to multiple ply cracks that tend to be more localized and, thus, more detrimental in thinner specimens.

V-Notch (Iosipescu) Shear Test (ASTM standard D 5379). The V-notch (Iosipescu) test method shown in Fig.

20 is based on modifications of the original test (Ref 34) by the composites group at the University of

Wyoming (Ref 35, 36). Analysis of the specimen under compressive loading reveals that a state of relatively

uniform shear stress exists in the center of the notched specimen (Ref 35, 36, 37, 38, 39). Because of the small

specimen size and opportunity for interlaminar shear testing, the test has become popular. Figure 21 shows the

specimen configuration for measuring shear response of unidirectional composites, but the test has also

successfully been used for laminated, woven fabric, and chopped fiber composites.

Fig. 20 Iosipescu shear test. (a) Testing configuration. (b) Specimen

Fig. 21 Iosipescu test specimen configurations and active shear stresses

The specimen (Fig. 20b) should be carefully machined. The loading faces should be ground smooth and parallel

to avoid out-of-plane bending and twisting. Good quality notches with rounded tips are required to minimize

stress concentrations at the notch tips and the occurrence of early failures, such as splitting in unidirectional, 0°

specimens. Specimen thickness is nominally 2.5 mm (0.1 in.). A ±45° strain gage rosette should be attached to

the center test section. If bending and twisting are of concern, the use of back-to-back strain gages has been

suggested (Ref 39).

The specimen should be carefully centered and clamped in the test fixture using the specimen alignment pin

(see Fig. 20a). Compressive load should be delivered by the crosshead at a rate of 1.5 mm/min (0.05 in./min)

while monitoring the load and strain signals until the specimen fails. Determination of the shear stress-strain

response of the material is based on shear stress calculated from:

= P/A

(Eq 11)

where A is the test-section cross-sectional area in the notched region of the specimen. For 0° specimens,

detailed stress analysis (Ref 39) has revealed a nonuniform stress state. Shear stress calculation incorporating a

correction factor is suggested below:

= 0.84P/A

(Eq 12)

for a specimen with 90° angle notches. After early initiation of notch-tip damage, however, the stress becomes

more uniform, and the use of the correction factor of 0.84 may be unwarranted (Ref 10). Shear strain, γ

, is

simply calculated from the ±45° strain gage readings as:

= |ε

| + |ε

-45

(Eq 13)

or, if only a 45° gage is used:

= 2|ε

(Eq 14)

Failure Modes. Damage in the form of longitudinal matrix cracks initiating from the notches is a common

occurrence in 0° unidirectional specimens. Load drops are observed when the cracks propagate, but the splits

eventually arrest. The specimen can carry further load until final failure occurs, which involves multiple matrix

cracks. 90° specimens fail early due to matrix cracking and do not produce a representative failure stress.

Cross-ply [0/90°] lay-ups show distributed microcracking, which is representative for laminated composites.

[±45°] and quasi-isotropic lay-ups are difficult to test because of the high loads required for failure.

Rail Shear Test. The rail shear test is described in ASTM D 4255. The in-plane shear properties of

unidirectional, laminated, and discontinuous fiber composites are determined by imposing axial load on a

fixture consisting of two pairs of tensile loaded rails or three pairs of compressive loaded rails (see Fig. 22).

Usually the specimen is bolted to the rails, but slippage is common at increasing loads for some composites.

Adhesive bonding of the specimen to the rails has been suggested (Ref 40). Figure 23 shows the specimen

geometries for the two-rail and three-rail tests. The diameter of the holes in the specimens exceeds that of the

fasteners so that load is only transferred through shear. The specimens should be instrumented with ±45° strain

gage rosettes as indicated in Fig. 22 and 23. To alleviate stress concentrations and cracking, especially in 0°

unidirectional specimens at free edges at loaded corners, different specimen designs have been suggested (Ref

40), but such cracking may not be critical in 90° and [0/90°] lay-ups. The specimen thicknesses should be

between 1.3 and 3.2 mm (0.05 and 0.13 in.) to avoid shear buckling while failing at reasonable loads.

Tolerances for specimen and fixture dimensions are not provided in the ASTM D 4255 guide, but the

dimensions are listed within 0.1 mm (0.04 in.), which should indicate the necessary accuracy.

Fig. 22 Rail shear test. (a) Two-rail configuration. (b) Three-rail configuration

Fig. 23 Specimen geometries for rail shear tests. (a) Two-rail test. (b) Three-rail test. All dimensions are

in millimeters.

The test specimen should be bolted to the test fixture using a torque wrench (see ASTM D 4255). The

crosshead speed should be 1.5 mm/min. (0.06 in./min), and load and strains should be monitored until the

specimen fails. The shear stress is calculated from:

= P/bh (two rail)

(Eq 15a)

= P/2bh (three rail)

(Eq 15b)

where b and h are specimen length and thickness, respectively. The shear strain is determined from:

= |ε

| + |ε

-45

(Eq 16)

A stress-strain curve can thus be constructed, which enables shear modulus, G

, to be determined from the

initial slope of the curve and the shear strength, S

, to be determined from the maximum stress.

Failure Modes. Due to stress concentrations at the corners formed at the intersection of the steel rails and the

specimen, damage is likely to initiate at these locations. Such damage is usually in the form of matrix cracks

that propagate along the fibers in a unidirectional ply. In a 0° lay-up, such cracks cause a large scatter in shear

strength (Ref 40, 41). Tests with the fibers perpendicular to the rails or cross-ply [0/90°] laminates produce

many ply cracks, but such cracking may not be detrimental and is often considered acceptable. Hussain and

Adams (Ref 42) have developed a modification of the two-rail shear test that eliminates the need for holes, thus

improving the stress distribution in the specimen. The simplification reduces the cost of fabrication

significantly while producing results consistent with expectations from other tests. Tests on quasi-isotropic and

[±45°] lay-ups using the standard method have revealed delamination, and the results are not conclusive. The

Wyoming-modified method (Ref 42) produced much better results on the [±45°] lay-ups.

Torsion Tube Test. The torsion tube test (Fig. 24), is especially suited for in-plane shear testing of filament-

wound structural parts, and has been recognized as an excellent shear test (Table 3) considered for ASTM

standardization. The specimen, however, is difficult to fabricate from prepreg and can obviously not be used to

test materials in the form of flat sheets. The specimen is bonded to the end fixtures using potting compound,

instrumented with at least two three-element strain gage rosettes (-45/0/45°) with the reference direction being

the tube axis (Ref 43). The gage length-to-inner diameter ratio is unity (see Fig. 24), and wall thickness-to-

diameter ratio is 0.02. This small ratio leads to minor variations in the shear stress across the wall, and the tube

may be considered as “thin-walled.”

Fig. 24 Torsion tube shear test. All dimensions are in millimeters.

The shear stress is calculated from:

= Tr/I

(Eq 17)

where T is the torque applied, r is the mean radius, and I

is the polar moment of inertia. For a thin-walled tube,

is given by (Ref 28):

= 2πr

(Eq 18)

where t is the wall thickness. Equation (17) then becomes:

= T/2πr

(Eq 19)

Shear strain, γ

, is evaluated from the gages at 45° and -45°:

= |ε

| + |ε

-45

(Eq 20)

An axial gage may be used to verify the absence of axial strains, that is, for a good shear test, ε(0°) ~ 0.

Failure Modes. The tube specimen configuration under torque load produces a pure state of shear stress over

most of the test specimen volume except near the grips. Hoop wound tubes fail perpendicular to the tube axis,

away from the grips at low stress levels. Failure stresses and strains in tubes with axial fiber orientation or

cross-ply [0/90°] lay-ups are larger than those for 90° tubes and comparable to other tests, but such tubes are

difficult to make. Buckling is a possible failure mode for thin-walled tubes when the shear strength is large (Ref

44), such as in tubes containing ±45° angle plies. Such failures invalidate the shear test.

Footnote

The section “Interlaminar Shear Properties of Fiber-

Reinforced Composites at High Strain Rates” was written

by John Harding and Stephen Hallett, Oxford University. The section “Fatigue Testing and Behavior of Fiber-

Reinforced Composites” was written by W. Steven Johnson and Ramesh Talreja, Georgia Institute of

Technology.

References cited in this section

4. J.M. Whitney, I.M. Daniel, and R.B. Pipes, Experimental Mechanics of Fiber Reinforced Composite

Materials, revised ed., Society for Experimental Mechanics; Prentice-Hall, Englewood Cliffs, 1984

5. R.B. Pipes, Test Methods, Delaware Composites Design Encyclopedia, Vol 6, (L.A. Carlsson and J.W.

Gillespie, Jr., Ed.), Technomic, Lancaster, 1990, p 3

8. L.A. Carlsson and R.B. Pipes, Experimental Characterization of Advanced Composite Materials, 2nd

ed., Technomic, Lancaster, 1987

9. D.W. Wilson and L.A. Carlsson, Mechanical Characterization of Composite Materials, Physical

Methods of Chemistry, 2nd. ed., Vol VII, B.W. Rossiter and R.B. Baetzold, Ed., John Wiley & Sons,

1991, p 139

10. S. Chatterjee, D.F. Adams, and D.W. Oplinger, “Test Methods for Composites - A Status Report,” (Vol

I, Tension Test Methods; Vol II, Compression Test Methods; Vol III. Shear Test Methods),

DOT/FAA/CT-93/17-I, II, III, FAA Technical Center, Atlantic City International Airport, NJ, June 1993

11. D.W. Oplinger, K.R. Gandhi, and B.S. Parker, “Studies of Tension Test Specimens for Composite

Material Testing,” AMMRC TR 82-27, Army Materials and Mechanics Research Center, Watertown,

MA, 1982

12. M.H. Kural and D.L. Flaggs, A Finite Element Analysis of Composite Tension Specimens, Compos.

Technol. Rev., Vol 3, 1983, p 11

13. M.G. Abdallah and R.L. Westberg, Effect of Tab Design of the ASTM D 3039 Tension Specimen on

Delivered Strength of HMS 501-6 Graphite/Epoxy Materials, Proc. 1987 SEM Spring Conf. on

Experimental Mechanics, 1987, p 362

14. Product Catalog, Wyoming Test Fixtures, Laramie, Wyoming, 1992

15. L.J. Hart-Smith, “Some Observations About Test Specimens and Structural Analysis for Fibrous

Composites,” ASTM STP 1059, 1988, p 86

16. W.S. Johnson, S.J. Lubowinski, and A.L. Highsmith, Mechanical Characterization of Unnotched

SCS/Ti-15-3 Metal S

Matrix Composites at Room Temperature, ASTM STP 1080, 1990, p 193

17. R. Walsh, private communication

18. K.F. Port, “The Compressive Strength of Carbon Fiber Reinforced Plastics,” Technical Report 82083,

Royal Aircraft Establishment, Farnborough, England, Aug 1982

19. D.F. Adams and E.Q. Lewis, Influence of Specimen Gage Length and Loading Method on the Axial

Compressive Strength of a Unidirectional Composite Material, Exp. Mech., Vol 31, 1991, p 14

20. E.M. Odom and D.F. Adams, Failure Modes of Unidirectional Carbon/Epoxy Composite Specimens,

Composites, Vol 21, 1990, p 289

21. K.E. Hofer, Jr. and P.N. Rao, A New Static Testing Compression Fixture for Advanced Composite

Materials, J. Testing Eval., Vol 5, 1977, p 278

22. M.N. Irion and D.F. Adams, Compression Creep Testing of Unidirectional Composites, Composites,

Vol 12, 1981, p 117

23. “Advanced Composite Compression Tests”, Boeing Specification Support Standard BSS 7260, The

Boeing Company, Seattle, WA, Dec 1988

24. L.J. Hart-Smith, Backing Out Equivalent Unidirectional Strengths from Tests on Cross-Plied Laminates,

Proc. 3rd International SAMPE Symposium and Exhibition, 1992, p 977

25. D.W. Wilson, V. Altstadt, M. Maier, J. Prandy, K. Thoma, and D. Vinckier, An Analytical and

Experimental Evaluation of 0/90 Laminate Tests for Compression Characterization, J. Compos.

Technol. Res., Vol 16 (No. 2), Apr 1994, p 146–153

26. J.S. Welsh and D.F. Adams, “Unidirectional Composite Compression Strengths Obtained by Cross-Ply

Laminates,” J. Compos. Technol. Res., Vol 18 (No. 4), Oct 1996, p 241

27. D.F. Adams and J.S. Welsh, The Wyoming Combined Loading Compression (CLC) Test Method, J.

Compos. Technol. and Res., Vol 19 (No. 3), 1997, p 123

28. J.M. Gere and S.P. Timoshenko, Mechanics of Materials, 2nd. ed., PWS Eng., Boston, 1984

29. J.M. Whitney and C.E. Browning, On Short-Beam Shear Tests for Composite Materials, Exp. Mech.,

Vol 25, 1985, p 294

30. J.M. Whitney, Structural Analysis of Laminated Anisotropic Plates, Technomic, Lancaster, 1987

31. B.W. Rosen, A Simple Procedure for Experimental Determination of the Longitudinal Shear Modulus

of Unidirectional Composites, J. Compos. Mater., Vol 6, 1972, p 552

32. S. Kellas and J. Morton, “Scaling Effects in Angle-Ply Laminates,” NASA CR 4423, Feb 1992

33. S. Kellas, J. Morton, and K. E. Jackson, An Evaluation of the ±45 Tensile Test for the Determination of

the In-Plane Shear Strength of Composite Materials, Proc. of the ICCM-8 Conf., Honolulu, July 1991

34. N. Iosipescu, New Accurate Procedure for Single Shear Testing of Metals, J. Mater., Vol 2, 1967, p 537

35. D.E. Walrath and D.F. Adams, The Iosipescu Shear Test as Applied to Composite Materials, Exp.

Mech., March 1983, p 105–110

36. D.E. Walrath and D.F. Adams, “Verification and Application of the Iosipescu Shear Test Method,”

UWME-DR-401-103-1, University of Wyoming, Laramie, June 1984

37. M.G. Abdallah, D.S. Gardener, and H.E. Gascoigne, An Evaluation of Graphite/Epoxy Iosipescu Shear

Specimen Testing Methods with Optical Techniques, Proc. of the 1985 SEM Spring Conf. on

Experimental Mechanics, Society of Experimental Mechanics, June 1985, p 833

38. H. Ho, M.Y. Tsai, J. Morton, and G.L. Farley, “An Evaluation of the Iosipescu Specimen for Composite

Materials Shear Property Measurement,” VPI & SU Report CCMS-91-18, Center for Composite

Materials and Structures, Aug 1991

39. S.N. Chatterjee, E.C.J. Wung, C.F. Yen, V. Ramnath, J.A. Kessler and D.F. Adams, “Composite

Specimen Design Analysis,” MTL Report TR-91-5, Vol I and II, 1991

40. R. Garcia, T.A. Weisshar, and R.R. McWhitney, “An Experimental and Analytical Investigation of the

Rail Shear Test Method as Applied to Composite Materials,” Exp. Mech., Vol 20, 1980, p 273

41. R.J. Butler, P.M. Barnard, and P.T. Curtis, The Development of a Satisfactory, Simple, Shear Fatigue

Test for Unidirectional E-Glass/Epoxy, ASTM STP 972, 1988, p 227

42. A. Hussain and D.F. Adams, “The Wyoming-Modified Two-Rail Shear Test Fixture for Composite

Materials,” J. Compos. Technol. Res., Vol 21 (No. 4), 1999, p 215

43. R. Bucinell, “Interim JANNAF Test Method for In-Plane Shear Properties of Unidirectional Fiber/Resin

Composite Cylinders,” Interim Report for MIL-HDBK-17, AMTL Contract DAAL04-89-C-0023, 1991

44. G.A. Foley, M.E. Roylance, and W.W. Houghton, “Use of Torsion Tubes to Measure In-Plane Shear

Properties of Filament Wound Composites,” STP 1003, ASTM 1989, p 208

Mechanical Testing of Fiber-Reinforced Composites

Dale Wilson, The Johns Hopkins University, Leif A. Carlsson, Florida Atlantic University

Structural Laminate Properties

The lamina and laminate tests fail to address the effect of holes or cutouts in laminates, a critical requirement

for structural design. Also, designers need experimental data quantifying the effect of impact induced damage

on laminates. Of particular concern is the compression strength of impact damaged laminates, since

sublaminates formed by delamination are susceptible to buckling that can produce failures at substantially

lower failure loads than undamaged laminates. Open hole tension, open hole compression, and compression

after impact tests have been developed to characterize the behavior of composite laminates with cutouts or

damage. Methods have been written for these tests but none are currently standardized. Data from the open hole

tests is routinely used in design analysis. The compression after impact test is used primarily in screening

candidate materials.

Open Hole Tension Test

Cutouts and holes are requirements in many structural applications. The effect of cutouts in composite

laminates is greater than the effect caused by the reduction in load-carrying area alone. Stress concentrations

are produced in the laminate adjacent to cutout boundaries that substantially reduce load-carrying capacity.

Stress concentrations are a function of laminate anisotropy and cutout geometry. Sharp notches produce higher

stress concentration factors than circular cutouts. However, the notch sensitivity of laminates is significantly

influenced by laminate stacking sequence and a host of microstructural materials characteristics like matrix

toughness, matrix stiffness, and fiber to matrix adhesion. High stress concentrations produce complex damage

zones, which in turn redistribute the stress and increase the energy required to produce failure significantly

above that predicted from the stress concentration factor alone. It has been shown that larger notches produce

lower strengths, because the stress concentrations involve a larger volume, increasing the probability of failure

due to a critical flaw (Ref 45). The stress distribution illustrated in Fig. 25 is the basis for the point stress

criterion for notched strength prediction (Ref 46), which states that failure occurs when the stress at some

characteristic distance d

reaches the unnotched tensile strength of the composite.

Fig. 25 Stress concentration adjacent to a hole in a composite laminate subjected to uniaxial loading

The test method for open hole tension uses a circular cutout in a test specimen (Fig. 26). The method is now

standardized as test method ASTM D 5766. It employs a 305 mm long by 38 mm wide (12 in. by 1.5 in.)

specimen containing a 6.35 mm (0.25 in.) hole. Quasi-isotropic laminate configurations are specified to be

(+45/0/-45/90°)

for tape or (±45/(0/90°))

for fabric prepregs. While other laminate configurations and

geometries are possible, it is recommended that the width-to-hole diameter ratio of 6 be maintained.

Fig. 26 Open hole tension test specimen geometry. All dimensions are in millimeters.

The specimen should be machined to the specification shown in Fig. 26. Tolerance on the hole location relative

to the specimen centerline is critical, since eccentricity can significantly decrease strengths. Specimens can be

tabbed or untabbed, although untabbed specimens reduce cost. If ultimate strain and modulus are desired,

specimens may be instrumented with a strain gage located on the specimen centerline 25 mm (1 in.) from the

hole center. ASTM D 5766, however, covers only notched strength and does not contain provisions for strain

measurement.

The test is performed as a uniaxial tension test following ASTM D 3039. The specimen is loaded until tension

failure occurs through the notch. If failure occurs outside the notch, the test result should be discarded, since the

failure was caused by a flaw in the material. If failures consistently fall outside the notch area, the naturally

occurring flaws in the material are larger than the notch (this is possible with some sheet molding compounds).

Then, the specimen design must be scaled to reflect the material inhomogeneity level. At least five specimens

should be tested per test condition.

The notched strength σ

is calculated as the tensile strength of the laminate based on the far-field stress:

= P/bd

(Eq 21)

where P is the maximum load, b is the specimen width, and d is the specimen thickness. If the specimen is

instrumented, the modulus is determined as:

(Eq 22)

where P

and P

are the loads at 1000 and 3000 microstrain, respectively. The strain at failure is determined

from the stress-strain curve.

Notched strength data is typically used for materials screening and for determining design allowables. For

design, it is necessary to generate empirical data based on the material, the laminate configuration, and the hole

sizes required. In lieu of generating empirical data for every conceivable material, laminate, and hole size

combination, it is possible to use the point stress criterion (PSC) analysis to interpolate notched strength over a

range of hole diameters by testing a series of three different notch sizes for the material and laminate

construction of interest (Ref 8, 47).

Open Hole Compression Test

Aside from the added complexity associated with specimen loading, much of what was described for the open

hole tension test applies to the open hole compression test. Damage develops differently, and the failure mode

reflects sensitivity to the microbuckling failure mechanisms that contribute to compression failure in composite

materials. Gynn and Bradley (Ref 48) studied the damage mechanisms in open hole compression specimens and

found that matrix stiffness plays a significant role in the development of a damage zone and the type of fiber

failure mechanism that leads to ultimate failure. In ductile matrix composites, fiber kinking leads to the

formation of a shear crippling zone across the net section of the hole, eventually leading to catastrophic failure.

Brittle systems fail quickly and catastrophically, without evidence of the shear crippling behavior.

The open hole compression strength can be measured using the same specimen as the open hole tension test.

One difference (shown in Fig. 27) is that more stringent tolerances are specified for the machining of the

specimen ends. For end loaded compression tests the perpendicularity of the specimen ends with the loading

direction is critical. The quality of the hole and the location of the hole relative to the loading centerline are also

important. The compression test commonly employs the SACMA fixture shown in Fig. 28. The specimen is

face supported during testing (see Fig. 28). The fixture incorporates a cutout around the specimen hole so as not

to constrain damage propagation and artificially increase the failure load. It also provides a milled out section

located 25 mm (1 in.) from the specimen cutout to accommodate a strain gage.