《纺织复合材料》课程参考文献（Composite Materials Handbook，Volume 1）CHAPTER 6 LAMINA, LAMINATE, AND SPECIAL FORM CHARACTERIZATION

MIL-HDBK-17-1F Volume 1,Chapter 6 Lamina,Laminate,and Special Form Characterization to within 0.001 millivolt,which is the precision to which the NIST thermocouple reference tables are pub- lished. If a self-scaling output device is used,its precision(and verified accuracy)must be sufficient to pro- vide at least four times the accuracy desired from the calibration. The Ice Point Reference The ice point reference is used to bring the T2 junction to its standard value of 32F(0C)during cali- bration of the probe.Again,this is only necessary if either the calibration standard or the probe being calibrated does not have another form of cold junction compensation.Ice point reference chambers are often simply a well-controlled and monitored ice bath.Electronic ice points are also available which greatly simplify the setup.It is important to note that the wiring from the T2 end of the thermocouple (which is at 32F(0C)in the ice point reference)to the readout device should be exclusively copper. This ensures that the emf response of the assembly is as-assumed by the thermocouple reference tables. 6.4.6 Data acquisition systems This section is reserved for future use. 6.5 TESTING ENVIRONMENTS This section is reserved for future use. 6.5.1 Introduction This section is reserved for future use. 6.5.2 Laboratory ambient test environment This section is reserved for future use. 6.5.3 Non-ambient testing environment 6.5.3.1 Introduction Composite materials can be affected by exposure to non-laboratory ambient environmental conditions and so must be tested to determine those effects.Below laboratory ambient conditions as well as above laboratory ambient conditions must be included in the test matrix to determine each effects.Guidelines for the above and below laboratory ambient test conditions are included below.Many different regimes of testing may be appropriate depending on the usage of the materials.Normal environmental conditions for terrestrial applications would be from as cold as-67F(-55C)and up to 350F(180C).Conditions in space would widen the band of performance interest from-250F to 450F(-160C to 230C).Cryogenic conditions(less than-250F(-160C))may be of interest for storage tank usage.Special conditions may dictate the usage of composite materials up to and beyond the short duration limit of 600F (315C) around leading edges or engine components.The user must determine what the limits for their particular application may be to allow for proper non-laboratory ambient testing to be completed on the materials used in the application. The purpose of this section is to give the user some guidance in the testing of materials under other than standard laboratory conditions.Both below and above room temperature test conditions are dis- cussed below.Further guidance related to non-laboratory ambient testing can be received from SACMA SRM 11R-94,Recommended Method for Environmental Conditioning of Composite Laminates. 6-26

MIL-HDBK-17-1F Volume 1, Chapter 6 Lamina, Laminate, and Special Form Characterization 6-26 to within 0.001 millivolt, which is the precision to which the NIST thermocouple reference tables are pub￾lished. If a self-scaling output device is used, its precision (and verified accuracy) must be sufficient to pro￾vide at least four times the accuracy desired from the calibration. The Ice Point Reference The ice point reference is used to bring the T2 junction to its standard value of 32°F (0°C) during cali￾bration of the probe. Again, this is only necessary if either the calibration standard or the probe being calibrated does not have another form of cold junction compensation. Ice point reference chambers are often simply a well-controlled and monitored ice bath. Electronic ice points are also available which greatly simplify the setup. It is important to note that the wiring from the T2 end of the thermocouple (which is at 32°F (0°C) in the ice point reference) to the readout device should be exclusively copper. This ensures that the emf response of the assembly is as-assumed by the thermocouple reference tables. 6.4.6 Data acquisition systems This section is reserved for future use. 6.5 TESTING ENVIRONMENTS This section is reserved for future use. 6.5.1 Introduction This section is reserved for future use. 6.5.2 Laboratory ambient test environment This section is reserved for future use. 6.5.3 Non-ambient testing environment 6.5.3.1 Introduction Composite materials can be affected by exposure to non-laboratory ambient environmental conditions and so must be tested to determine those effects. Below laboratory ambient conditions as well as above laboratory ambient conditions must be included in the test matrix to determine each effects. Guidelines for the above and below laboratory ambient test conditions are included below. Many different regimes of testing may be appropriate depending on the usage of the materials. Normal environmental conditions for terrestrial applications would be from as cold as -67°F (-55°C) and up to 350°F (180°C). Conditions in space would widen the band of performance interest from -250°F to 450°F (-160°C to 230°C). Cryogenic conditions (less than -250°F (-160°C)) may be of interest for storage tank usage. Special conditions may dictate the usage of composite materials up to and beyond the short duration limit of 600°F (315°C) around leading edges or engine components. The user must determine what the limits for their particular application may be to allow for proper non-laboratory ambient testing to be completed on the materials used in the application. The purpose of this section is to give the user some guidance in the testing of materials under other than standard laboratory conditions. Both below and above room temperature test conditions are dis￾cussed below. Further guidance related to non-laboratory ambient testing can be received from SACMA SRM 11R-94, Recommended Method for Environmental Conditioning of Composite Laminates

MIL-HDBK-17-1F Volume 1,Chapter 6 Lamina,Laminate,and Special Form Characterization 6.5.3.2 Subambient testing Testing performed at below laboratory ambient test temperatures can present unique challenges. Special fixturing or lubrication may be needed to ensure that properties measured are material behavior related and not due to freezing or sticking of sliding surfaces.Materials can become more brittle and change their failure modes.Special instrumentation may be necessary to record material properties at the colder temperatures.Adhesives used for tabbing or strain gaging should be types that retain their elongation at the cold temperatures. Test temperatures as cold as-67F(-55C)are common and are discussed here.The test setup in a test chamber must be precooled until stabilized at test temperature.Fixturing should be allowed to stabi- lize prior to testing.Cooling medium may be liquid nitrogen(LN2),liquid carbon dioxide(LCO2),or a re- frigerated chamber.Temperature measurements are commonly made with J,K or T type thermocouples (T/C's).See Section 6.4.5 for more information on temperature measurement.A dummy test specimen should be used to determine soak times prior to actual testing.The dummy specimen should be fabri- cated using the same material and ply orientation as the test specimen.To determine the soak time,a T/C should be inserted into a hole drilled at the centerline of the dummy specimen.Record the time it takes to reach the desired test temperature.This time should be used when testing to regulate when the test specimens are at the appropriate test temperature.Cool down rates should be controlled to mini- mize thermal shock and possibility of damage and/or microcracking. Freezing of test fixtures can be a cause of anomalous test results.Fixture clearances must be checked to ensure free sliding surfaces exist.Proper lubricants or no lubricants should be used at the cold temperatures to prevent any fixture related effects on the test results. A thermocouple(T/C)should be placed in contact with the surface of the test specimen at the time of test.A typical soak time of 5-10 minutes,or the time determined from actual experimentation,should be used,after reaching test temperature.Appropriate safety equipment should always be worn to prevent cold burns.Care must be taken if using LN2 or LCO2 when cooling the chamber to ensure that room oxy- gen is not depleted. 6.5.3.3 Above ambient testing Testing performed at above ambient temperatures must be done with consideration for the tempera- ture and moisture content of the test sample.Special fixturing may be needed to accommodate the high temperatures.The possibility for adhesive failures and drying of test specimens should be evaluated be- fore proceeding with a test program.Special lubricants may be required to prevent fixturing from sticking or binding.Instrumentation made especially for the required temperatures must be used to ensure valid data is recorded.Strain gages,extensometers,and adhesives with the correct temperature rating must be identified and used.Special strain gage foils or backing materials may be required to withstand the elevated temperatures during testing.Instrumentation may require additional calibration at test tempera- tures. The above ambient test temperatures,to 350F(180C),are discussed here.The test setup in a test chamber must be heated until stabilized at test temperature.Fixturing should be allowed to stabilize prior to testing.Heating of the test fixture with specimen or only the specimen is usually accomplished with an electrically heated chamber.Temperature measurements are commonly made using J.K,or T type ther- mocouples(T/C's).A dummy test specimen should be used to determine soak times prior to actual test- ing.The dummy specimen should be fabricated using the same material and ply orientation as the test specimen.To determine the soak time,a T/C should be inserted into a hole drilled at the centerline of the dummy specimen.Record the time it takes to reach the desired test temperature.This time should be used when testing to regulate when the test specimens are at the appropriate test temperature.Heat up rates should be controlled to minimize thermal shock and possibility of damage and/or microcracking. 6-27

MIL-HDBK-17-1F Volume 1, Chapter 6 Lamina, Laminate, and Special Form Characterization 6-27 6.5.3.2 Subambient testing Testing performed at below laboratory ambient test temperatures can present unique challenges. Special fixturing or lubrication may be needed to ensure that properties measured are material behavior related and not due to freezing or sticking of sliding surfaces. Materials can become more brittle and change their failure modes. Special instrumentation may be necessary to record material properties at the colder temperatures. Adhesives used for tabbing or strain gaging should be types that retain their elongation at the cold temperatures. Test temperatures as cold as -67°F (-55°C) are common and are discussed here. The test setup in a test chamber must be precooled until stabilized at test temperature. Fixturing should be allowed to stabi￾lize prior to testing. Cooling medium may be liquid nitrogen (LN2), liquid carbon dioxide (LCO2), or a re￾frigerated chamber. Temperature measurements are commonly made with J, K or T type thermocouples (T/C’s). See Section 6.4.5 for more information on temperature measurement. A dummy test specimen should be used to determine soak times prior to actual testing. The dummy specimen should be fabri￾cated using the same material and ply orientation as the test specimen. To determine the soak time, a T/C should be inserted into a hole drilled at the centerline of the dummy specimen. Record the time it takes to reach the desired test temperature. This time should be used when testing to regulate when the test specimens are at the appropriate test temperature . Cool down rates should be controlled to mini￾mize thermal shock and possibility of damage and/or microcracking. Freezing of test fixtures can be a cause of anomalous test results. Fixture clearances must be checked to ensure free sliding surfaces exist. Proper lubricants or no lubricants should be used at the cold temperatures to prevent any fixture related effects on the test results. A thermocouple (T/C) should be placed in contact with the surface of the test specimen at the time of test. A typical soak time of 5-10 minutes, or the time determined from actual experimentation, should be used, after reaching test temperature. Appropriate safety equipment should always be worn to prevent cold burns. Care must be taken if using LN2 or LCO2 when cooling the chamber to ensure that room oxy￾gen is not depleted. 6.5.3.3 Above ambient testing Testing performed at above ambient temperatures must be done with consideration for the tempera￾ture and moisture content of the test sample. Special fixturing may be needed to accommodate the high temperatures. The possibility for adhesive failures and drying of test specimens should be evaluated be￾fore proceeding with a test program. Special lubricants may be required to prevent fixturing from sticking or binding. Instrumentation made especially for the required temperatures must be used to ensure valid data is recorded. Strain gages, extensometers, and adhesives with the correct temperature rating must be identified and used. Special strain gage foils or backing materials may be required to withstand the elevated temperatures during testing. Instrumentation may require additional calibration at test tempera￾tures. The above ambient test temperatures, to 350°F (180°C), are discussed here. The test setup in a test chamber must be heated until stabilized at test temperature. Fixturing should be allowed to stabilize prior to testing. Heating of the test fixture with specimen or only the specimen is usually accomplished with an electrically heated chamber. Temperature measurements are commonly made using J, K, or T type ther￾mocouples (T/C’s). A dummy test specimen should be used to determine soak times prior to actual test￾ing. The dummy specimen should be fabricated using the same material and ply orientation as the test specimen. To determine the soak time, a T/C should be inserted into a hole drilled at the centerline of the dummy specimen. Record the time it takes to reach the desired test temperature. This time should be used when testing to regulate when the test specimens are at the appropriate test temperature. Heat up rates should be controlled to minimize thermal shock and possibility of damage and/or microcracking

MIL-HDBK-17-1F Volume 1,Chapter 6 Lamina,Laminate,and Special Form Characterization Excessive heat up rates may cause charring or melting of test specimens or adhesives.An appropri- ate lubricant,such as molybdenum disulfide,should be used on sliding surfaces to ensure freedom of movement of test fixtures. A T/C should be placed in contact with the surface of the test specimen prior to testing.A standard soak time would be 5-10 minutes,after reaching test temperature,if the test condition is dry.A standard soak time would be 2 minutes,after reaching test temperature,if the test condition is wet,to prevent too much dryout of the test specimen. If moisture content is a testing variable,then the dryout of the test specimen,unless humidity is con- trolled during test,should be evaluated by weighing a traveler before and after a specimen soak time and test.See Section 6.3 for moisture conditioning guidelines.Appropriate safety equipment should always be worn to prevent burns. For moderate test conditions,i.e.,less than 200F(93C),a humidity controlled test chamber is op- tional for short duration tests.When testing above 200F(93C),then a precise humidity control is im- practical and specimen dryout is a concern,especially for fatigue testing.Soak times prior to test should be kept short(<3 min.to minimize the dryout. Testing performed at temperatures above 350F(180C)must use special strain gages and strain gage adhesives,extensometry,and fixturing designed for the elevated temperatures.Special high tem- perature capable tab materials and tab adhesives will need to be utilized to prevent tab failures.Usage of these materials may be inappropriate at other temperatures. Thermocouples are the most common transducer for measuring temperatures.Various T/C types may be used but J,K,and T are the most common.Some special conditions may dictate the use of re- sistance temperature detectors (RTD's).See Section 6.4.5 for more information on temperature meas- urement. 6.6 THERMAL/PHYSICAL PROPERTY TESTS The physical analysis methods for laminae and laminates provide information on the integrity of the fabricated composite.Thermal analysis methods are used to determine the glass transition and crystal- line melt temperatures,coefficient of thermal expansion,and residual heat of reaction.Additional analyti- cal methods discussed in the following sections are used to determine fiber volume,void volume,density, dimensional stability,and moisture weight gain. 6.6.1 Introduction The thermal analytical techniques described in Chapter 4.Section 4.5.2 may also be used to evaluate composite materials.Information obtained from thermal analysis includes the glass transition tempera- ture,crystalline melt temperature,expansion/contraction properties,thermal stability.and extent of cure for thermosets. 6.6.2 Extent of cure Characterization of extent of cure of composite materials has become increasingly important as con- trolled staging of complex or thick parts has been implemented as part of advanced processing schemes. Debulking and staging of stiffeners or other structural details can be used to facilitate assembly of large complex parts,with the ultimate goal of allowing out-of-autoclave processing.Debulking and staging are also a critically important aspect of the fabrication of thick parts to prevent resin migration and fiber wavi- ness. Several different thermal analysis techniques are commonly used for extent of cure measurements in fiber reinforced organic matrix composites.There include differential scanning calorimetry(DSC)or dy- 6-28

MIL-HDBK-17-1F Volume 1, Chapter 6 Lamina, Laminate, and Special Form Characterization 6-28 Excessive heat up rates may cause charring or melting of test specimens or adhesives. An appropri￾ate lubricant, such as molybdenum disulfide, should be used on sliding surfaces to ensure freedom of movement of test fixtures. A T/C should be placed in contact with the surface of the test specimen prior to testing. A standard soak time would be 5-10 minutes, after reaching test temperature, if the test condition is dry. A standard soak time would be 2 minutes, after reaching test temperature, if the test condition is wet, to prevent too much dryout of the test specimen. If moisture content is a testing variable, then the dryout of the test specimen, unless humidity is con￾trolled during test, should be evaluated by weighing a traveler before and after a specimen soak time and test. See Section 6.3 for moisture conditioning guidelines. Appropriate safety equipment should always be worn to prevent burns. For moderate test conditions, i.e., less than 200°F (93°C), a humidity controlled test chamber is op￾tional for short duration tests. When testing above 200°F (93°C), then a precise humidity control is im￾practical and specimen dryout is a concern, especially for fatigue testing. Soak times prior to test should be kept short (<3 min.) to minimize the dryout. Testing performed at temperatures above 350°F (180°C) must use special strain gages and strain gage adhesives, extensometry, and fixturing designed for the elevated temperatures. Special high tem￾perature capable tab materials and tab adhesives will need to be utilized to prevent tab failures. Usage of these materials may be inappropriate at other temperatures. Thermocouples are the most common transducer for measuring temperatures. Various T/C types may be used but J, K, and T are the most common. Some special conditions may dictate the use of re￾sistance temperature detectors (RTD’s). See Section 6.4.5 for more information on temperature meas￾urement. 6.6 THERMAL/PHYSICAL PROPERTY TESTS The physical analysis methods for laminae and laminates provide information on the integrity of the fabricated composite. Thermal analysis methods are used to determine the glass transition and crystal￾line melt temperatures, coefficient of thermal expansion, and residual heat of reaction. Additional analyti￾cal methods discussed in the following sections are used to determine fiber volume, void volume, density, dimensional stability, and moisture weight gain. 6.6.1 Introduction The thermal analytical techniques described in Chapter 4, Section 4.5.2 may also be used to evaluate composite materials. Information obtained from thermal analysis includes the glass transition tempera￾ture, crystalline melt temperature, expansion/contraction properties, thermal stability, and extent of cure for thermosets. 6.6.2 Extent of cure Characterization of extent of cure of composite materials has become increasingly important as con￾trolled staging of complex or thick parts has been implemented as part of advanced processing schemes. Debulking and staging of stiffeners or other structural details can be used to facilitate assembly of large complex parts, with the ultimate goal of allowing out-of-autoclave processing. Debulking and staging are also a critically important aspect of the fabrication of thick parts to prevent resin migration and fiber wavi￾ness. Several different thermal analysis techniques are commonly used for extent of cure measurements in fiber reinforced organic matrix composites. There include differential scanning calorimetry (DSC) or dy-

MIL-HDBK-17-1F Volume 1,Chapter 6 Lamina,Laminate,and Special Form Characterization namic thermal analysis(DTA)to measure the extent of the residual curing exotherm and dynamic me- chanical analysis(DMA)or thermomechanical analysis (TMA)to measure the glass transition tempera- ture.Measurement of Tg is discussed in some detail in Section 6.6.3 below. 6.6.3 Glass transition temperature 6.6.3.1 Overview The glass transition of a polymer matrix composite is a temperature-induced change in the matrix ma- terial from the glassy to the rubbery state during heating,or from a rubber to a glass during cooling.A change in matrix stiffness of two to three orders of magnitude occurs during the glass transition,due to the onset or freezing out of long range molecular mobility of the polymer chains.The temperature at which the glass transition occurs is a function of the molecular architecture and crosslink density of the polymer chains,but it is also dependent on the heating or cooling rate used in the measurement,and on test frequency if a dynamic mechanical technique is employed.In addition to the change in stiffness,the glass transition is marked by a change in the heat capacity and the coefficient of thermal expansion of the material,and so has at least some characteristics of a second order thermodynamic transition(see Ref- erence6.6.3.1). The glass transition is frequently characterized by a glass transition temperature(Tg),but since the transition often occurs over a broad temperature range,the use of a single temperature to characterize it may give rise to some confusion.The experimental technique used to obtain the T must be described in detail,especially temperature scanning rate and frequency used.The method by which T.is calculated from the data must also be clearly stated.Reported T may reflect onset of the glass transition or mid- point temperature depending on the data reduction method. Upon exposure to high humidity environments,polymer matrices will absorb environmental moisture and be plasticized by it.One effect of this plasticization is the depression of T.frequently by a significant amount.A highly crosslinked resin (one based for instance on a tetrafunctional epoxide such as TGMDA) may have a high initial T.but it may be depressed more strongly than that in a less highly crosslinked system.Measurement of the T.in a composite material plasticized by absorbed moisture poses some difficult experimental challenges.Heating the test specimen as required by the measurement will drive off at least some of the absorbed moisture,thereby affecting the measured properties. Due to the decrease in matrix stiffness that occurs at the glass transition and to the low strength of these polymer matrices in the rubbery state,the matrix can no longer function effectively to transfer load to the fibers or suppress fiber buckling above the glass transition.T is,therefore,frequently used to de- fine the upper use temperature of a composite material,although the time-dependent properties of the material such as creep compliance may be more sensitive to temperature within the glass transition range than are the quasi-static mechanical properties.A safety margin of 50F(28C)between the T and the material operational limit(MOL)has been proposed for epoxy matrix composites (see Section 2.2.8). This approach is useful for initially estimating the MOL,or for verifying a previously chosen MOL.How- ever,since glass transition frequently occurs over a temperature range,and the measured value of T is highly dependent on method,supplemental mechanical property tests should be considered,particularly for new material systems(see Section 2.2.8). 6.6.3.2 T,Measurements Several different methods have been used to characterize the glass transition in polymeric materials, and most of these are also applicable to fiber reinforced materials. 6.6.3.2.1 Differential scanning calorimetry(DSC) Since the heat capacity of a composite material changes at the glass transition,differential scanning calorimetry(DSC)may be used to determine T.The glass transition is detected as a shift in the heat flow versus temperature curve (see Figure 6.6.3.2.1).Many calorimeters are supplied with software 6-29

MIL-HDBK-17-1F Volume 1, Chapter 6 Lamina, Laminate, and Special Form Characterization 6-29 namic thermal analysis (DTA) to measure the extent of the residual curing exotherm and dynamic me￾chanical analysis (DMA) or thermomechanical analysis (TMA) to measure the glass transition tempera￾ture. Measurement of Tg is discussed in some detail in Section 6.6.3 below. 6.6.3 Glass transition temperature 6.6.3.1 Overview The glass transition of a polymer matrix composite is a temperature-induced change in the matrix ma￾terial from the glassy to the rubbery state during heating, or from a rubber to a glass during cooling. A change in matrix stiffness of two to three orders of magnitude occurs during the glass transition, due to the onset or freezing out of long range molecular mobility of the polymer chains. The temperature at which the glass transition occurs is a function of the molecular architecture and crosslink density of the polymer chains, but it is also dependent on the heating or cooling rate used in the measurement, and on test frequency if a dynamic mechanical technique is employed. In addition to the change in stiffness, the glass transition is marked by a change in the heat capacity and the coefficient of thermal expansion of the material, and so has at least some characteristics of a second order thermodynamic transition (see Ref￾erence 6.6.3.1). The glass transition is frequently characterized by a glass transition temperature (Tg), but since the transition often occurs over a broad temperature range, the use of a single temperature to characterize it may give rise to some confusion. The experimental technique used to obtain the Tg must be described in detail, especially temperature scanning rate and frequency used. The method by which Tg is calculated from the data must also be clearly stated. Reported Tg may reflect onset of the glass transition or mid￾point temperature depending on the data reduction method. Upon exposure to high humidity environments, polymer matrices will absorb environmental moisture and be plasticized by it. One effect of this plasticization is the depression of Tg, frequently by a significant amount. A highly crosslinked resin (one based for instance on a tetrafunctional epoxide such as TGMDA) may have a high initial Tg, but it may be depressed more strongly than that in a less highly crosslinked system. Measurement of the Tg in a composite material plasticized by absorbed moisture poses some difficult experimental challenges. Heating the test specimen as required by the measurement will drive off at least some of the absorbed moisture, thereby affecting the measured properties. Due to the decrease in matrix stiffness that occurs at the glass transition and to the low strength of these polymer matrices in the rubbery state, the matrix can no longer function effectively to transfer load to the fibers or suppress fiber buckling above the glass transition. Tg is, therefore, frequently used to de￾fine the upper use temperature of a composite material, although the time-dependent properties of the material such as creep compliance may be more sensitive to temperature within the glass transition range than are the quasi-static mechanical properties. A safety margin of 50F° (28C°) between the Tg and the material operational limit (MOL) has been proposed for epoxy matrix composites (see Section 2.2.8). This approach is useful for initially estimating the MOL, or for verifying a previously chosen MOL. How￾ever, since glass transition frequently occurs over a temperature range, and the measured value of Tg is highly dependent on method, supplemental mechanical property tests should be considered, particularly for new material systems (see Section 2.2.8). 6.6.3.2 Tg Measurements Several different methods have been used to characterize the glass transition in polymeric materials, and most of these are also applicable to fiber reinforced materials. 6.6.3.2.1 Differential scanning calorimetry (DSC) Since the heat capacity of a composite material changes at the glass transition, differential scanning calorimetry (DSC) may be used to determine Tg. The glass transition is detected as a shift in the heat flow versus temperature curve (see Figure 6.6.3.2.1). Many calorimeters are supplied with software

MIL-HDBK-17-1F Volume 1,Chapter 6 Lamina,Laminate,and Special Form Characterization which may be used to calculated T.T:of neat resin specimens is relatively easy to detect with DSC,but in composite specimens the resin content in the specimen is small,and the more highly crosslinked the resin,the smaller the change in heat capacity.It is,therefore,sometimes difficult to detect T in highly crosslinked cured composites (see Reference 6.6.3.2.1). MOI Temperature FIGURE 6.6.3.2.1 Differential scanning calorimetry(DSC). 6.6.3.2.2 Thermomechanical analysis(TMA) Thermomechanical techniques such as expansion,flexure,or penetration thermomechanical analysis (TMA)may also be used to determine T.In expansion TMA,the coefficient of thermal expansion a is measured as a function of temperature.As noted above,a undergoes a change during the glass transi- tion,and T is determined by the point of intersection of lines fit to the thermal expansion data above and below the glass transition range.Figure 6.6.3.2.2 illustrates the specimen geometries and data reduction methods used for various TMA techniques. In flexural TMA,a rectangular specimen is loaded in bending and the dimensional change is meas- ured as a function of temperature.A curve fitting technique as illustrated in Figure 6.6.3.2.2 is used to calculate T.Flexural TMA measurement of T is similar to heat distortion temperature(HDT)measure- ment,since in both cases the specimens are loaded in flexure.An HDT specimen may be a full-size flex- ural test specimen,and is loaded in three-point bending or as a cantilever beam.Displacement is meas- ured as a function of temperature,and the HDT is the temperature at which the displacement reaches some predetermined value.Use of a full-size specimen minimizes moisture loss during the HDT test,but flexural TMA and HDT measurement share the disadvantage that values of T or HDT obtained will be sensitive to the modulus of the reinforcing fibers in the composite sample and they will give different re- sults depending on the nature of the fiber. As shown in Figure 6.6.3.2.2,penetration mode TMA measures the hardness of the material.One disadvantage of this technique is that if the probe is touching a reinforcing fiber,an accurate measure- ment of the T of the matrix will not be obtained. 6-30

MIL-HDBK-17-1F Volume 1, Chapter 6 Lamina, Laminate, and Special Form Characterization 6-30 which may be used to calculated Tg. Tg of neat resin specimens is relatively easy to detect with DSC, but in composite specimens the resin content in the specimen is small, and the more highly crosslinked the resin, the smaller the change in heat capacity. It is, therefore, sometimes difficult to detect Tg in highly crosslinked cured composites (see Reference 6.6.3.2.1). FIGURE 6.6.3.2.1 Differential scanning calorimetry (DSC). 6.6.3.2.2 Thermomechanical analysis (TMA) Thermomechanical techniques such as expansion, flexure, or penetration thermomechanical analysis (TMA) may also be used to determine Tg. In expansion TMA, the coefficient of thermal expansion α is measured as a function of temperature. As noted above, α undergoes a change during the glass transi￾tion, and Tg is determined by the point of intersection of lines fit to the thermal expansion data above and below the glass transition range. Figure 6.6.3.2.2 illustrates the specimen geometries and data reduction methods used for various TMA techniques. In flexural TMA, a rectangular specimen is loaded in bending and the dimensional change is meas￾ured as a function of temperature. A curve fitting technique as illustrated in Figure 6.6.3.2.2 is used to calculate Tg. Flexural TMA measurement of Tg is similar to heat distortion temperature (HDT) measure￾ment, since in both cases the specimens are loaded in flexure. An HDT specimen may be a full-size flex￾ural test specimen, and is loaded in three-point bending or as a cantilever beam. Displacement is meas￾ured as a function of temperature, and the HDT is the temperature at which the displacement reaches some predetermined value. Use of a full-size specimen minimizes moisture loss during the HDT test, but flexural TMA and HDT measurement share the disadvantage that values of Tg or HDT obtained will be sensitive to the modulus of the reinforcing fibers in the composite sample and they will give different re￾sults depending on the nature of the fiber. As shown in Figure 6.6.3.2.2, penetration mode TMA measures the hardness of the material. One disadvantage of this technique is that if the probe is touching a reinforcing fiber, an accurate measure￾ment of the Tg of the matrix will not be obtained