7 Mechanical Property Measurement 7.1 Introduction Mechanical testing of materials and structural details is conducted to satisfy one or more of the following objectives:1)characterization of materials or processes,2)development of design allowables,3)qualification of materials for certain applications,4)quality control,5)assessment of strength and durability under sustained or cyclic loads,or 6)assessment of the influence of damage or degradation on residual strength. Aerospace wrought metal alloys are available in standard pre-fabricated forms with well-characterized properties.By contrast,composites are usually formed at the same time as the component is manufactured and therefore can have a very wide range of properties depending on the fiber,resin,lay-up,volume fraction, etc.Some properties of composites are more sensitive to environmental conditions.Thus,testing requirements are generally more demanding than is the case for metals. The use of mechanical testing for developing design allowables for composites is described in Chapter 13,and its use in the testing of adhesively bonded or mechanically fastened joints is described in Chapter 9. 7.1.1 Types of Mechanical Tests Most tests are conducted under static tensile,compressive,or shear loading. They may also be conducted under flexural loading,which induce tensile, compressive,and shear stresses in the various zones of the specimen.The static loading may be of short duration,taking only a few minutes,as in a standard tensile test to measure strength or stiffness. Static tests,most generally performed under tensile loading,may be also be prolonged for weeks or months,as in a creep test,to measure the long-term strength and strain stability-often at elevated temperature.These tests are usually conducted at various percentages of the short-term ultimate tensile strength,typically 10-50%. Cyclic loading tests to measure resistance to degradation and cracking under varying loads are essentially repeated static loading.The frequency of application is generally low,in the case of composites around 5-10 Hz,to avoid heating.Loading may be tension/tension,tension/compression,or reversed 213
7 Mechanical Property Measurement 7.1 Introduction Mechanical testing of materials and structural details is conducted to satisfy one or more of the following objectives: 1) characterization of materials or processes, 2) development of design allowables, 3) qualification of materials for certain applications, 4) quality control, 5) assessment of strength and durability under sustained or cyclic loads, or 6) assessment of the influence of damage or degradation on residual strength. Aerospace wrought metal alloys are available in standard pre-fabricated forms with well-characterized properties. By contrast, composites are usually formed at the same time as the component is manufactured and therefore can have a very wide range of properties depending on the fiber, resin, lay-up, volume fraction, etc. Some properties of composites are more sensitive to environmental conditions. Thus, testing requirements are generally more demanding than is the case for metals. The use of mechanical testing for developing design allowables for composites is described in Chapter 13, and its use in the testing of adhesively bonded or mechanically fastened joints is described in Chapter 9. 7.1.1 Types of Mechanical Tests Most tests are conducted under static tensile, compressive, or shear loading. They may also be conducted under flexural loading, which induce tensile, compressive, and shear stresses in the various zones of the specimen. The static loading may be of short duration, taking only a few minutes, as in a standard tensile test to measure strength or stiffness. Static tests, most generally performed under tensile loading, may be also be prolonged for weeks or months, as in a creep test, to measure the long-term strength and strain stability-----often at elevated temperature. These tests are usually conducted at various percentages of the short-term ultimate tensile strength, typically 10-50%. Cyclic loading tests to measure resistance to degradation and cracking under varying loads are essentially repeated static loading. 1 The frequency of application is generally low, in the case of composites around 5-10 Hz, to avoid heating. Loading may be tension/tension, tension/compression, or reversed 213
214 COMPOSITE MATERIALS FOR AIRCRAFT STRUCTURES shear at constant amplitude or under spectrum loading,and may be aimed at simulating the actual loading conditions in a particular application. Dynamic loads are used to measure the resistance of the materials to impact or ballistic conditions.These tests are also conducted under tension,compression, shear,or flexure,or they may be conducted using an impactor or penetrator of some type.In some tests,the impact event may occur while the specimen is under tensile or compressive loading.Typically,loading occurs over a 1-millisecond time interval. Testing may be conducted at different temperatures and levels of absorbed moisture.They may also include exposure to a range of other environmental conditions,such as UV and solvents. The specimens may be simple coupons or they may be structural details with representative stress-raisers such as holes,filled with a fastener or open.The coupons or details may include representative damage such as sharp notches or impact damage. Test machines consist of loading frames,one fixed and one moving crosshead, separated either by a simple electromechanical screw action or by a servo- hydraulic piston.For simple static testing,the screw-driven machines are simpler and less costly and there is less danger of overload caused by accidental rapid movement of the crosshead.However,for versatility in loading (e.g.spectrum loading in fatigue testing)and in load capacity,the servo-hydraulic machines are unmatched. 7.1.2 Special Requirements for Testing Composites During the early development of composites,many of the test techniques used for metals and other homogenous,isotropic materials were used to determine the properties of composite materials.It was soon recognized that anisotropic composite materials often required special consideration in terms of mechanical property determination.Much of the test method development was also undertaken within individual organizations,thus standardization was difficult and many methods developed were not adequate for the newer,emerging materials. Since those early days,there has been a great deal of effort devoted to the standardization of test methods,and there are a number of reference sources that can be used to identify the relevant techniques.Test standards have been published by the American Society for Testing and Materials(ASTM)2-and the Suppliers of Advanced Composite Materials Association (SACMA),10.11 together with other information sources such as the U.S.Department of Defense Military Handbook 17 (MIL-HDBK 17;Polymer Matrix Composites).MIL- HDBK 17 is specifically suited to composite materials for aerospace applications and is generally used for test method selection. The test techniques briefly described here are the ones most commonly used when measuring stress and strain in the tensile,compressive,flexural,and shear load states,but they are not the only techniques that can be used.The most critical
214 COMPOSITE MATERIALS FOR AIRCRAFT STRUCTURES shear at constant amplitude or under spectrum loading, and may be aimed at simulating the actual loading conditions in a particular application. Dynamic loads are used to measure the resistance of the materials to impact or ballistic conditions. These tests are also conducted under tension, compression, shear, or flexure, or they may be conducted using an impactor or penetrator of some type. In some tests, the impact event may occur while the specimen is under tensile or compressive loading. Typically, loading occurs over a 1-millisecond time interval. Testing may be conducted at different temperatures and levels of absorbed moisture. They may also include exposure to a range of other environmental conditions, such as UV and solvents. The specimens may be simple coupons or they may be structural details with representative stress-raisers such as holes, filled with a fastener or open. The coupons or details may include representative damage such as sharp notches or impact damage. Test machines consist of loading frames, one fixed and one moving crosshead, separated either by a simple electromechanical screw action or by a servohydraulic piston. For simple static testing, the screw-driven machines are simpler and less costly and there is less danger of overload caused by accidental rapid movement of the crosshead. However, for versatility in loading (e.g. spectrum loading in fatigue testing) and in load capacity, the servo-hydraulic machines are unmatched. 7.1.2 Special Requirements for Testing Composites During the early development of composites, many of the test techniques used for metals and other homogenous, isotropic materials were used to determine the properties of composite materials. It was soon recognized that anisotropic composite materials often required special consideration in terms of mechanical property determination. Much of the test method development was also undertaken within individual organizations, thus standardization was difficult and many methods developed were not adequate for the newer, emerging materials. Since those early days, there has been a great deal of effort devoted to the standardization of test methods, and there are a number of reference sources that can be used to identify the relevant techniques. Test standards have been published by the American Society for Testing and Materials (ASTM) 2-9 and the Suppliers of Advanced Composite Materials Association (SACMA), 1°'ix together with other information sources such as the U.S. Department of Defense Military Handbook 17 (MIL-HDBK 17; Polymer Matrix Composites). MILHDBK 17 is specifically suited to composite materials for aerospace applications and is generally used for test method selection. The test techniques briefly described here are the ones most commonly used when measuring stress and strain in the tensile, compressive, flexural, and shear load states, but they are not the only techniques that can be used. The most critical
MECHANICAL PROPERTY MEASUREMENT 215 issues that must be satisfied are that the test method used accurately creates the required stress state in the material and that the specimen failure be consistent with this stress state and not be artificially influenced by the test method. Because of the variabilties encountered in coupon testing,airworthiness authorities require multiple tests across several batches.MIL-HDBK 17 recommends a minimum of six specimens per test point and five batches of material to be tested.These requirements mean that the exploration of even a minimum of material properties entails a very large number of test specimens. When conducting tests to determine the strength and stiffness of a composite material,the first question that must be answered is"What mode of its performance is to be measured?"Composites,as with other materials,can have significantly different mechanical properties when tested in different ways.The main loading modes that are generally of interest are tension, compression,flexure,and shear-each has its own particular test techniques and difficulties. To facilitate design computations,the elastic properties of the composite lamina discussed in Chapter 6 are usually obtained first through simple coupon tests.Recall the relationships for in-plane elastic properties,noting that in most cases,in-plane properties will be used to design the laminate: 1/E1 -v12/E2 0 6 82 -v12/E1 1/E2 0 02 Y12 0 0 1/G12 T12 And since v2/E=v21/E2(See Chapter 6),only three tests are required to establish the in-plane elastic properties,in other words,0 tension,90 tension, and in-plane shear. Because it is not possible to conduct tests on single plies,the coupons are laid up with multiple plies,all orientated in the same direction.The exception is the in-plane shear in which,if a 45 tension test is selected (see below),plies are alternated between+and-45 symmetric about the center line. Strength values should not,in general,be taken from these coupons although they are very often taken to failure.The reasons for this are explained in Chapter 12.Laminate strength should be obtained from tests on representative laminates in which the orientations of the fiber lay-up are similar to those anticipated in the design.These values that are used in initial design are generally substantiated by tests on structural elements and often finally on full-scale structures.This is often referred to as the Testing Pyramid, which is illustrated in Figure 7.1. It must be understood that issues such as scale effects2 and complex load conditions13 can become important when testing composite components,and the data obtained from simple coupon tests can often only be used for comparing materials and not as accurate predictions of component behavior
MECHANICAL PROPERTY MEASUREMENT 215 issues that must be satisfied are that the test method used accurately creates the required stress state in the material and that the specimen failure be consistent with this stress state and not be artificially influenced by the test method. Because of the variabilties encountered in coupon testing, airworthiness authorities require multiple tests across several batches. MIL-HDBK 17 recommends a minimum of six specimens per test point and five batches of material to be tested. These requirements mean that the exploration of even a minimum of material properties entails a very large number of test specimens. When conducting tests to determine the strength and stiffness of a composite material, the first question that must be answered is "What mode of its performance is to be measured?" Composites, as with other materials, can have significantly different mechanical properties when tested in different ways. The main loading modes that are generally of interest are tension, compression, flexure, and shear--each has its own particular test techniques and difficulties. To facilitate design computations, the elastic properties of the composite lamina discussed in Chapter 6 are usually obtained first through simple coupon tests. Recall the relationships for in-plane elastic properties, noting that in most cases, in-plane properties will be used to design the laminate: 81 ~2 T12 l/El = --Vl2/E1 0 -lJ12/E 2 0 1/E2 0 0 l/G12 o" 1 0"2 '7"12 And since 1)12/E 1 m_ 1,21/E 2 (See Chapter 6), only three tests are required to establish the in-plane elastic properties, in other words, 0 ° tension, 90 ° tension, and in-plane shear. Because it is not possible to conduct tests on single plies, the coupons are laid up with multiple plies, all orientated in the same direction. The exception is the in-plane shear in which, if a 45 ° tension test is selected (see below), plies are alternated between + and - 45 ° symmetric about the center line. Strength values should not, in general, be taken from these coupons although they are very often taken to failure. The reasons for this are explained in Chapter 12. Laminate strength should be obtained from tests on representative laminates in which the orientations of the fiber lay-up are similar to those anticipated in the design. These values that are used in initial design are generally substantiated by tests on structural elements and often finally on full-scale structures. This is often referred to as the Testing Pyramid, which is illustrated in Figure 7.1. It must be understood that issues such as scale effects 12 and complex load conditions 13 can become important when testing composite components, and the data obtained from simple coupon tests can often only be used for comparing materials and not as accurate predictions of component behavior
216 COMPOSITE MATERIALS FOR AIRCRAFT STRUCTURES SUB-COMPONENTS DETAILS ELEMENTS Fig.7.1 Testing pyramid for composite structures. 7.2 Coupon Tests 7.2.1 Tension Valid axial tension testing,particularly of strong unidirectional composites, can be a challenge.The load must be transferred from the testing apparatus into the specimen via shear,and the shear strengths of composites are often significantly lower than their tensile strength.Thus shear failure within the gripping region is often a problem. The standard test technique (outlined in Refs.2 and 3 for open-hole tension) describes the use of a parallel-sided,rectangular specimen with bonded end tabs. However,these tabs,which are normally made from a glass fabric/epoxy composite,are not strictly required.The key factor is the successful introduction of load into the specimen.Therefore,if acceptable failures are being obtained with reasonable consistency,then it can be assumed that the gripping method is working.A wide variety of bonded tab or unbonded shim configurations have been used successfully.One unbonded shim material sometimes used is a coarse mesh made of carborundum-coated cloth. Load measurement is performed via the load cell in the test machine,and strain measurement is done by an extensometer secured to the specimen or by adhesively bonded strain gauges.To measure Poisson's ratio,both the axial and transverse strain must be measured concurrently.Extensometers are normally preferred because they are reusable,easier to mount,and often more reliable at elevated temperatures or in high-moisture-content environments.If strain gauges are used,then the active gauge length (length of specimen over which the strain is measured)is recommended to be at least 6 mm for tape composites and at least as large as the characteristic repeating unit of the weave for woven materials
216 COMPOSITE MATERIALS FOR AIRCRAFT STRUCTURES i SUtB_CnMpnNE~ITS /", ~ \'\ I \ L__~ ~_ ........... ~ ........ / cnup_~ ~ ',, ./,,,.,.__.._.,,_,_ ...,,_,._,..-_ ~_ _~_~__ ~_ ~_'.~ Fig. 7.1 Testing pyramid for composite structures. 7.2 Coupon Tests 7.2.1 Tension Valid axial tension testing, particularly of strong unidirectional composites, can be a challenge. The load must be transferred from the testing apparatus into the specimen via shear, and the shear strengths of composites are often significantly lower than their tensile strength. Thus shear failure within the gripping region is often a problem. The standard test technique (outlined in Refs. 2 and 3 for open-hole tension) describes the use of a parallel-sided, rectangular specimen with bonded end tabs. However, these tabs, which are normally made from a glass fabric/epoxy composite, are not strictly required. The key factor is the successful introduction of load into the specimen. Therefore, if acceptable failures are being obtained with reasonable consistency, then it can be assumed that the gripping method is working. A wide variety of bonded tab or unbonded shim configurations have been used successfully. One unbonded shim material sometimes used is a coarse mesh made of carborundum-coated cloth. Load measurement is performed via the load cell in the test machine, and strain measurement is done by an extensometer secured to the specimen or by adhesively bonded strain gauges. To measure Poisson's ratio, both the axial and transverse strain must be measured concurrently. Extensometers are normally preferred because they are reusable, easier to mount, and often more reliable at elevated temperatures or in high-moisture-content environments. If strain gauges are used, then the active gauge length (length of specimen over which the strain is measured) is recommended to be at least 6 mm for tape composites and at least as large as the characteristic repeating unit of the weave for woven materials
MECHANICAL PROPERTY MEASUREMENT 217 A successful test must cause failure within the gauge region.Failure at the tab edge(or gripped edge)or within the tab is unacceptable.Failure due to early edge delamination,which is normally caused by poor machining,is also unacceptable. Figure 7.2 illustrates typically a)unacceptable and b)acceptable specimen failures.Poor load system alignment is often a major contributor to premature failure,and it is highly desirable to evaluate system alignment with a suitably strain-gauged,alignment coupon 7.2.2 Compression There is still a great amount of debate among researchers as to the most appropriate method for compression testing or indeed whether there is a true axial compression test for composites.4.15 Generally,compression failure occurs through buckling,ranging from classical column buckling of the entire specimen Fig.7.2 Failure modes in tensile testing:a)unacceptable;b)acceptable
MECHANICAL PROPERTY MEASUREMENT 217 A successful test must cause failure within the gauge region. Failure at the tab edge (or gripped edge) or within the tab is unacceptable. Failure due to early edge delamination, which is normally caused by poor machining, is also unacceptable. Figure 7.2 illustrates typically a) unacceptable and b) acceptable specimen failures. Poor load system alignment is often a major contributor to premature failure, and it is highly desirable to evaluate system alignment with a suitably strain-gauged, alignment coupon. 7.2.2 Compression There is still a great amount of debate among researchers as to the most appropriate method for compression testing or indeed whether there is a true axial compression test for composites. 14'15 Generally, compression failure occurs through buckling, ranging from classical column buckling of the entire specimen a) 7 Fig. 7.2 Failure modes in tensile testing: a) unacceptable; b) acceptable
218 COMPOSITE MATERIALS FOR AIRCRAFT STRUCTURES cross-section to local microbuckling of fibers that often leads to failure through the process of kink band formation."Therefore,the greater resistance to buckling the test fixture provides to the specimen,the higher the value of compressive strength that is obtained. Many different test methods and specimen configurations have been developed over the decades in an attempt to limit specimen buckling,and there are a number of tests that have become the most widely used in current practice.The Illinois Institute of Technology Research Institute (IITRD method,4 which has become an ASTM standard,and the modified ASTM D695 method10 (currently a SACMA standard),are two methods used for un-notched specimens (Figs.7.3 and 7.4,respectively).The SACMA Recommended Test Method 3R-9411 is commonly used for open-hole compression testing Fig.7.5. As with the tension test,tabs are not absolutely required for the specimen, although they are strongly recommended for specimens made with unidirectional reinforcement.The main criteria is that correct failure occurs within the gauge LINEAR BEARINGS TABBED KNURLED SPECIMEN SURFACE CLAMPING CLAMPING WEDGES SCREWS LOAD- ALIGNMENT ALIGNMENT BLOCK RODS Fig.7.3 The IITRI compression test rig
218 COMPOSITE MATERIALS FOR AIRCRAFT STRUCTURES cross-section to local microbuckling of fibers that often leads to failure through the process of kink band formation. 4 Therefore, the greater resistance to buckling the test fixture provides to the specimen, the higher the value of compressive strength that is obtained. Many different test methods and specimen configurations have been developed over the decades in an attempt to limit specimen buckling, and there are a number of tests that have become the most widely used in current practice. The Illinois Institute of Technology Research Institute (IITRI) method, 4 which has become an ASTM standard, and the modified ASTM D695 method ]° (currently a SACMA standard), are two methods used for un-notched specimens (Figs. 7.3 and 7.4, respectively). The SACMA Recommended Test Method 3R-9411 is commonly used for open-hole compression testing Fig. 7.5. As with the tension test, tabs are not absolutely required for the specimen, although they are strongly recommended for specimens made with unidirectional reinforcement. The main criteria is that correct failure occurs within the gauge I ~ ~ .~"~. KNURLED WEDGES LINEAR TAOBED • ~i CLAMPING --SCREWS Lo, o ¸ ALIGNMENT ALIGNMENT Fig. 7.3 The IITRI compression test rig
MECHANICAL PROPERTY MEASUREMENT 219 COVER PLATE SPECIMEN PLACED BETWEEN COVER LOADING PLATE AND APPLIED SUPPORT PLATE THROUGH END LOADING OF TABBED SPECIMEN SUPPORT PLATE HARDENED INSERT BASE SUPPORT Fig.7.4 Modified ASTM D695 compression test rig. FRONT SUPPORT PLATE BACK SUPPORT PLATE TEST WINDOW FRONT SUPPORT ALIGNMENT PLATE PLATE SPECIMEN PLACED BACK SUPPORT BETWEEN FRONT AND PLATE BACK SUPPORT PLATES AND END LOADED Fig.7.5 Rig for the SACMA recommended test method 3R-94 for open-hole compression testing
MECHANICAL PROPERTY MEASUREMENT 219 SOVER PLATE LOADING APPLIED THROUGH END LOADING OF TABBED SPECIMEN SPECIMEN PLACED BETWEEN COVER PLATE AND SUPPORT PLATE SUPPORT PLATE HARDENED INSERT BASE SUPPORT Fig. 7.4 Modified ASTM D695 compression test rig. FRONT SUPPORT PLATE ~ SPECIMEN PLACED BETWEEN FRONT AND BACK SUPPORT PLATES AND END LOADED TEST WINDOW PLATE Fig. 7.5 Rig for the SACMA recommended test method 3R-94 for open-hole compression testing
220 COMPOSITE MATERIALS FOR AIRCRAFT STRUCTURES area;if this does not occur correctly,then the data point should not be used. Figure 7.6 illustrates examples of acceptable and unacceptable failures,with any failure residing solely in the tabbed or gripped region being considered unacceptable.Due to the very short gauge length,it is likely that the failure location could be near the grip/tab termination region;this is still an acceptable failure. Compression tests are very sensitive to the flatness and parallelism of the specimens and/or tabs,and within the test specifications,the required tolerances are outlined.Gripping of the specimens and system misalignment are generally the biggest cause of data scatter,and of particular relevance to the test methods that use stabilizing lateral supports(SACMA SRM 1R-94 and 3R-94)is the issue of bolt torque.An over-torque of the bolts allows more of the applied load to be carried by the lateral supports through friction,thereby increasing the apparent compressive strength.Generally,the bolts are tightened up to a"finger-tight" level,a fairly arbitrary measurement;however studies'5 have shown that the torque should not exceed approximately 1 Nm. 7.2.3 Flexure A flexure test is,without doubt,one of the simplest types of tests to perform and thus has long been popular(Fig.7.7).The main difficulty is that it does not a) b) Fig.7.6 Failure modes in compression testing:a)unacceptable;b)acceptable
220 COMPOSITE MATERIALS FOR AIRCRAFT STRUCTURES area; if this does not occur correctly, then the data point should not be used. Figure 7.6 illustrates examples of acceptable and unacceptable failures, with any failure residing solely in the tabbed or gripped region being considered unacceptable. Due to the very short gauge length, it is likely that the failure location could be near the grip/tab termination region; this is still an acceptable failure. Compression tests are very sensitive to the flatness and parallelism of the specimens and/or tabs, and within the test specifications, the required tolerances are outlined. Gripping of the specimens and system misalignment are generally the biggest cause of data scatter, and of particular relevance to the test methods that use stabilizing lateral supports (SACMA SRM 1R-94 and 3R-94) is the issue of bolt torque. An over-torque of the bolts allows more of the applied load to be carried by the lateral supports through friction, thereby increasing the apparent compressive strength. Generally, the bolts are tightened up to a "finger-tight" level, a fairly arbitrary measurement; however studies 15 have shown that the torque should not exceed approximately 1 Nm. 7.2.3 Flexure A flexure test is, without doubt, one of the simplest types of tests to perform and thus has long been popular (Fig. 7.7). The main difficulty is that it does not a) b Fig. 7.6 Failure modes in compression testing: a) unacceptable; b) acceptable
MECHANICAL PROPERTY MEASUREMENT 221 FLEXURE THREE POINT BENDING 15 PLIES 8.36cm 42.50N】 1143cm 1.27Gm 14.6IN.】 0.60N.I 44 1.27cm 0.01N. 1.27cm FLEXURE:FOUR POINT BENDING 0.601N.) ↑.27cm (0.50N.J Fig.7.7 ASTM test methods for the flexure test. provide basic material property information because of the variation in stress- state within the specimen.The stress-state on the loading side is compression and on the supported side is tension;the mid-plane of the specimen is in pure shear. Therefore,depending on the relative values of the tension,compression,and shear strengths of the material,any one of these properties may be measured. The ratio of the support span length to specimen thickness is normally set long enough so that shear failure does not occur(32:1 is common)but whether failure initiates on the tensile or compressive face will be dependent on the material. Although the flexure test does not provide basic design data,its use is normally justified if the actual components are subjected to flexure.This is a valid argument if the span length to thickness ratio is similar to the laboratory test specimens.If not,the failure mode of the component in service may be different, and thus any comparison of the laboratory testing is not valid. The details of the standard flexure test are contained within the ASTM specification D790-84a'and this provides the recommended dimensions and cross-head speeds for various materials.There are two possible test configurations that can be used:three-point bending and four-point bending (Fig.7.7).Although the three-point bending test requires less material,the four- point bending has the advantage that uniform tensile or compressive stresses (with zero shear)are produced over the area between the loading points,not just under the loading point as in three-point bending test.In the three-point test,the high local stresses at the loading point affects the failure mode and load.It should be noted that excessive bending of the specimen before failure can render the test
MECHANICAL PROPERTY MEASUREMENT 221 FLEXURE" THREE POINt 8ENDING 15 PLIES ,27 em ~O,EO INJ Fig. 7.7 ASTM test methods for the flexure test. provide basic material property information because of the variation in stressstate within the specimen. The stress-state on the loading side is compression and on the supported side is tension; the mid-plane of the specimen is in pure shear. Therefore, depending on the relative values of the tension, compression, and shear strengths of the material, any one of these properties may be measured. The ratio of the support span length to specimen thickness is normally set long enough so that shear failure does not occur (32 : 1 is common) but whether failure initiates on the tensile or compressive face will be dependent on the material. Although the flexure test does not provide basic design data, its use is normally justified if the actual components are subjected to flexure. This is a valid argument if the span length to thickness ratio is similar to the laboratory test specimens. If not, the failure mode of the component in service may be different, and thus any comparison of the laboratory testing is not valid. The details of the standard flexure test are contained within the ASTM specification D790-84a 5 and this provides the recommended dimensions and cross-head speeds for various materials. There are two possible test configurations that can be used: three-point bending and four-point bending (Fig. 7.7). Although the three-point bending test requires less material, the fourpoint bending has the advantage that uniform tensile or compressive stresses (with zero shear) are produced over the area between the loading points, not just under the loading point as in three-point bending test. In the three-point test, the high local stresses at the loading point affects the failure mode and load. It should be noted that excessive bending of the specimen before failure can render the test
222 COMPOSITE MATERIALS FOR AIRCRAFT STRUCTURES invalid due to slipping of the specimen over the support points.This situation is discussed within the test specification. In specimens with sandwich construction having relatively thin skins on a honeycomb (or other suitable)core,loading in flexure simultaneously provides tensile stresses in one skin and compressive stresses in the other.This form of testing is particularly advantageous for compression testing of composites because a much larger area of skin can be tested than is possible in the standard tests,and the loading is far more realistic. 7.2.4 Shear Shear testing of composites is often a cause of confusion.Many different test procedures have been used,and only now are some gaining acceptance.This situation is hampered by the fact that many techniques cannot provide both shear strength and modulus from the one test. The ideal test for shear is torsion of a thin-walled tube,which provides a pure shear stress-state,yet this method is not often used.The specimens are relatively expensive,fragile,and difficult to hold and align correctly,and the technique requires a torsion-testing machine of sufficient capacity.Currently,the two-rail shear test(Fig.7.8)and the Iosipescu test'(Fig.7.9)are the most commonly used,although,it should be noted that the rail shear test is currently issued by ASTM as a Standard Guide,not a Standard Method.Both of these tests are not recommended for specimens containing only +45 fibers;rather,these specimens should be tested using the method outlined in Ref.8,which involves the use of a routine tensile test. Difficulties can arise when using the rail shear test because the specimens generally fail by out-of-plane buckling,therefore the measured values of strength and strain will be affected.Stress concentrations can also occur at the rail attachments,and suitable design of the rail area is critical to prevent failure occurring here.Due to these problems,shear data obtained using this test is often questioned. The losipescu test,nevertheless,is gaining in popularity due to having none of the disadvantages of the rail shear test and having the added advantages of using much less material and producing an essentially pure shear stress;however,shear stress concentrations develop at the root of the notches.Another advantage is that the test can be used to measure shear properties in any orientation.Thus,the losipescu method can be used to provide interlaminar shear properties by machining the specimen so that the interlaminar plane is parallel to the plane of the gauge area.The test specification contains recommendations for dimensions, but it is critical that the gauge area contain a sufficient number of fabric repeat units to ensure material properties are obtained.ASTM gives no guidelines on this but it is generally accepted that a minimum of 3 repeat units are used, therefore the specimen should be scaled up to achieve this.Twisting of the
222 COMPOSITE MATERIALS FOR AIRCRAFT STRUCTURES invalid due to slipping of the specimen over the support points. This situation is discussed within the test specification. In specimens with sandwich construction having relatively thin skins on a honeycomb (or other suitable) core, loading in flexure simultaneously provides tensile stresses in one skin and compressive stresses in the other. This form of testing is particularly advantageous for compression testing of composites because a much larger area of skin can be tested than is possible in the standard tests, and the loading is far more realistic. 7.2.4 Shear Shear testing of composites is often a cause of confusion. Many different test procedures have been used, and only now are some gaining acceptance. This situation is hampered by the fact that many techniques cannot provide both shear strength and modulus from the one test. The ideal test for shear is torsion of a thin-walled tube, which provides a pure shear stress-state, yet this method is not often used. The specimens are relatively expensive, fragile, and difficult to hold and align correctly, and the technique requires a torsion-testing machine of sufficient capacity. Currently, the two-rail shear test 6 (Fig. 7.8) and the Iosipescu test 7 (Fig. 7.9) are the most commonly used, although, it should be noted that the rail shear test is currently issued by ASTM as a Standard Guide, not a Standard Method. Both of these tests are not recommended for specimens containing only +45 ° fibers; rather, these specimens should be tested using the method outlined in Ref. 8, which involves the use of a routine tensile test. Difficulties can arise when using the rail shear test because the specimens generally fail by out-of-plane buckling, therefore the measured values of strength and strain will be affected. Stress concentrations can also occur at the rail attachments, and suitable design of the rail area is critical to prevent failure occurring here. Due to these problems, shear data obtained using this test is often questioned. The Iosipescu test, nevertheless, is gaining in popularity due to having none of the disadvantages of the rail shear test and having the added advantages of using much less material and producing an essentially pure shear stress; however, shear stress concentrations develop at the root of the notches. Another advantage is that the test can be used to measure shear properties in any orientation. Thus, the Iosipescu method can be used to provide interlaminar shear properties by machining the specimen so that the interlaminar plane is parallel to the plane of the gauge area. The test specification contains recommendations for dimensions, but it is critical that the gauge area contain a sufficient number of fabric repeat units to ensure material properties are obtained. ASTM gives no guidelines on this but it is generally accepted that a minimum of 3 repeat units are used, therefore the specimen should be scaled up to achieve this. Twisting of the