MIL-HDBK-17-1F Volume 1,Chapter 6 Lamina,Laminate,and Special Form Characterization Several styles of measuring faces are available.These generally fall into four categories:flat,spheri- cal,blade,and pointed.Both faces on a given instrument may be the same style or different(one face flat and one spherical,for example).Pointed faces are not recommended for use with composites,as they may penetrate the surface (pointed faces are typically used to measure the root diameter of threads).Blade (knife edge)faces are convenient for measuring specimen thickness between bonded tabs on short gage section length specimens.However,such specimens should be carefully inspected for the presence of tab bonding adhesive in the gage section.If adhesive is present the measured lami- nate thickness will be erroneously inflated. Flat and spherical(ball)faces are appropriate for most specimen width and thickness measurements, but laminate surface texture should be considered when choosing between these two face styles.For "glass smooth"surfaces,double flat,double ball,or ball-flat faces are all appropriate.However,if the sur- face is textured(due to coarse weave fabrics,or from use of peel ply during processing,as examples)a flat face will contact the "hills"of the texture,and the resulting measurement will be falsely inflated.A ball face,which will settle somewhat into the "valleys"of the texture or compress the "hills,"is therefore pre- ferred.Although the percentage error can vary with specific surface conditions,it is usually not significant for thick specimens.However,for thin(2-3 ply)specimens,measurements may be significantly biased since differences of 0.0015 to 0.0030 inch(0.038 to 0.076 mm)may typically be observed between measurements made with double ball and double flat micrometers.Test specimens that are smooth on one surface and textured on the other may be evaluated by a ball/flat micrometer. In addition to "stand-alone"micrometers,some testing machines have micrometers integrated into their systems,permitting direct electronic input of specimen dimensions.The system generally prompts the user to position the specimen in the micrometer for width,thickness,and possibly other measure- ments,and later uses these measurements for calculations.Since the measuring faces fall into the same categories as discussed above,the same considerations apply. 6.4.2.4 Scaled calipers Scaled calipers are devices with parallel,jaw-like measuring faces and a scale for reading the dis- tance between the stationary face and the movable face.Although models are available for measuring dimensions up to several feet,6 inch and 12 inch(15 cm and 30 cm)lengths are most common for meas- uring composite test specimens.The scale may be engraved along the length of the caliper,or may take the form of a dial or digital electronic readout.Although an engraved scale(with auxiliary vernier scale) and the digital readout have 0.0001 inch (2.5 um)resolution,accuracy is more commonly limited to ±0.001inch(0.025mm). Calipers are convenient for measuring specimen lengths and widths,particularly in the range of 1-12 inches(2.5-30 cm),since this range exceeds the capability of the common 1 inch micrometer.In addi- tion,some calipers have measuring tips (nibs)designed in such a way that internal as well as external measurements may be made.With this design,calipers may be used to measure hole diameters (in open hole tension and compression specimens,for example).Typically.nibs designed for internal measurement can fit into a 0.25 inch(6.35 mm)or larger hole.Some can read an internal dimension as small as 0.125 inch(3.18 mm). Calipers may not be particularly suited for measuring specimen thicknesses,especially if the speci- men surface(s)is textured.For such measurements a ball-faced instrument is generally preferred (see Section 6.4.2.3 above)as opposed to calipers(which have flat or blade shaped measuring faces). 6.4.2.5 Precision scales Precision scales are available in various lengths,with 6 inch and 12 inch(15 cm and 30 cm)being common.These tools are similar to rulers,but are usually made of steel and are more precisely and finely graduated.Each instrument typically has four scales,one along each edge of each side.The finest graduations are commonly 1/64 inch or 1/100(0.01)inch(0.4 mm or 0.25 mm).Reading to 1/100 inch 6-11
MIL-HDBK-17-1F Volume 1, Chapter 6 Lamina, Laminate, and Special Form Characterization 6-11 Several styles of measuring faces are available. These generally fall into four categories: flat, spherical, blade, and pointed. Both faces on a given instrument may be the same style or different (one face flat and one spherical, for example). Pointed faces are not recommended for use with composites, as they may penetrate the surface (pointed faces are typically used to measure the root diameter of threads). Blade (knife edge) faces are convenient for measuring specimen thickness between bonded tabs on short gage section length specimens. However, such specimens should be carefully inspected for the presence of tab bonding adhesive in the gage section. If adhesive is present the measured laminate thickness will be erroneously inflated. Flat and spherical (ball) faces are appropriate for most specimen width and thickness measurements, but laminate surface texture should be considered when choosing between these two face styles. For "glass smooth" surfaces, double flat, double ball, or ball-flat faces are all appropriate. However, if the surface is textured (due to coarse weave fabrics, or from use of peel ply during processing, as examples) a flat face will contact the "hills" of the texture, and the resulting measurement will be falsely inflated. A ball face, which will settle somewhat into the "valleys" of the texture or compress the "hills," is therefore preferred. Although the percentage error can vary with specific surface conditions, it is usually not significant for thick specimens. However, for thin (2-3 ply) specimens, measurements may be significantly biased since differences of 0.0015 to 0.0030 inch (0.038 to 0.076 mm) may typically be observed between measurements made with double ball and double flat micrometers. Test specimens that are smooth on one surface and textured on the other may be evaluated by a ball/flat micrometer. In addition to "stand-alone" micrometers, some testing machines have micrometers integrated into their systems, permitting direct electronic input of specimen dimensions. The system generally prompts the user to position the specimen in the micrometer for width, thickness, and possibly other measurements, and later uses these measurements for calculations. Since the measuring faces fall into the same categories as discussed above, the same considerations apply. 6.4.2.4 Scaled calipers Scaled calipers are devices with parallel, jaw-like measuring faces and a scale for reading the distance between the stationary face and the movable face. Although models are available for measuring dimensions up to several feet, 6 inch and 12 inch (15 cm and 30 cm) lengths are most common for measuring composite test specimens. The scale may be engraved along the length of the caliper, or may take the form of a dial or digital electronic readout. Although an engraved scale (with auxiliary vernier scale) and the digital readout have 0.0001 inch (2.5 µm) resolution, accuracy is more commonly limited to ±0.001 inch (0.025 mm). Calipers are convenient for measuring specimen lengths and widths, particularly in the range of 1 - 12 inches (2.5 - 30 cm), since this range exceeds the capability of the common 1 inch micrometer. In addition, some calipers have measuring tips (nibs) designed in such a way that internal as well as external measurements may be made. With this design, calipers may be used to measure hole diameters (in open hole tension and compression specimens, for example). Typically, nibs designed for internal measurement can fit into a 0.25 inch (6.35 mm) or larger hole. Some can read an internal dimension as small as 0.125 inch (3.18 mm). Calipers may not be particularly suited for measuring specimen thicknesses, especially if the specimen surface(s) is textured. For such measurements a ball-faced instrument is generally preferred (see Section 6.4.2.3 above) as opposed to calipers (which have flat or blade shaped measuring faces). 6.4.2.5 Precision scales Precision scales are available in various lengths, with 6 inch and 12 inch (15 cm and 30 cm) being common. These tools are similar to rulers, but are usually made of steel and are more precisely and finely graduated. Each instrument typically has four scales, one along each edge of each side. The finest graduations are commonly 1/64 inch or 1/100 (0.01) inch (0.4 mm or 0.25 mm). Reading to 1/100 inch
MIL-HDBK-17-1F Volume 1,Chapter 6 Lamina,Laminate,and Special Form Characterization (0.25 mm)generally requires use of a magnifying glass to discern the graduations clearly.While preci- sion scales may be used for any measurements requiring this resolution,calipers or other instruments are usually easier to use. 6.4.2.6 Rulers and tape measures These tools are commonly marked in 1/16 inch(1.6 mm)divisions,though some are marked in 1/32 inch(0.79 mm)increments for at least part of their total length.They are generally used for measure- ments that are recorded for descriptive purposes,but not for more precise measurements.For example, a ruler might be used to identify two groups of specimens:one group with four inch nominal gage lengths, and another group with six inch nominal gage lengths. 6.4.2.7 Special hole diameter measuring devices Although,as noted above in Section 6.4.2.4,some calipers are designed for inside diameter meas- urement,special instruments are also available for such measurements.These include telescoping gages,small hole gages,and calibrated pins.Telescoping gages are "T"shaped devices with two spring loaded plungers forming the top of the"T."The measuring faces at the plunger ends are curved and they self-center against the inside walls of the hole.Once in position they are locked by turning a knurled screw on the stem of the"T,"and the instrument is withdrawn from the hole.Hole diameter is then deter- mined by measuring the distance between the locked plunger faces using a standard micrometer.The disadvantages of telescoping gages are(1)a set of several gages must be used to cover a range of hole sizes and(2)because of their size,gages are not available for holes smaller than about 5/16 inch(8 mm) diameter. Small hole gages are similar to telescoping gages except that an adjustable split ball is used instead of plungers.The split ball is placed in the hole and is enlarged by turning the barrel of the device until the ball just contacts the hole walls.The instrument is then removed from the hole and measured with a standard micrometer in the same manner as the telescoping gage.These gages must also be used in sets to cover a wide range of hole sizes but,unlike the telescoping gages,holes down to about 1/8 inch diameter can be measured. Sets of calibrated pins of known diameter may also be used to measure hole diameters.Pins of vari- ous sizes are inserted into the hole until a close,but not tight,fit is obtained.The hole diameter is then taken as the pin diameter.Pins are available in virtually any size,and are generally graduated in 0.0005 inch increments.Very extensive sets are needed to cover a range of nominal hole sizes. Of the devices available,calipers or small hole gages are most useful and economical for measuring hole diameters in composite test specimens. 6.4.2.8 Calibration of dimensional measurement devices In order to maintain the stated accuracy of mechanical measuring devices such as micrometers and calipers,they must be periodically calibrated.In general,there are no detailed calibration procedures available in high level(ASTM,ANSI,etc.)U.S.standards.Typically these instruments are calibrated us- ing gage blocks,and specific procedures are contained in company internal specifications.Some ISO documents address aspects of this subject,and the reader is referred to standards under the jurisdiction of ISO Technical Committee 3 on Limits and Fits,as well as to ISO 10012-1(Reference 6.4.2.8). 6.4.3 Load measurement devices 6.4.3.1 Introduction The ability to accurately and repeatably measure load(force)is critical to the testing and characteri- zation of composite materials.This section will discuss the various types of instrumentation used to make load measurements,and provide guidelines to insure the accuracy of those measurements.Load meas- 6-12
MIL-HDBK-17-1F Volume 1, Chapter 6 Lamina, Laminate, and Special Form Characterization 6-12 (0.25 mm) generally requires use of a magnifying glass to discern the graduations clearly. While precision scales may be used for any measurements requiring this resolution, calipers or other instruments are usually easier to use. 6.4.2.6 Rulers and tape measures These tools are commonly marked in 1/16 inch (1.6 mm) divisions, though some are marked in 1/32 inch (0.79 mm) increments for at least part of their total length. They are generally used for measurements that are recorded for descriptive purposes, but not for more precise measurements. For example, a ruler might be used to identify two groups of specimens: one group with four inch nominal gage lengths, and another group with six inch nominal gage lengths. 6.4.2.7 Special hole diameter measuring devices Although, as noted above in Section 6.4.2.4, some calipers are designed for inside diameter measurement, special instruments are also available for such measurements. These include telescoping gages, small hole gages, and calibrated pins. Telescoping gages are "T" shaped devices with two spring loaded plungers forming the top of the "T." The measuring faces at the plunger ends are curved and they self-center against the inside walls of the hole. Once in position they are locked by turning a knurled screw on the stem of the "T," and the instrument is withdrawn from the hole. Hole diameter is then determined by measuring the distance between the locked plunger faces using a standard micrometer. The disadvantages of telescoping gages are (1) a set of several gages must be used to cover a range of hole sizes and (2) because of their size, gages are not available for holes smaller than about 5/16 inch (8 mm) diameter. Small hole gages are similar to telescoping gages except that an adjustable split ball is used instead of plungers. The split ball is placed in the hole and is enlarged by turning the barrel of the device until the ball just contacts the hole walls. The instrument is then removed from the hole and measured with a standard micrometer in the same manner as the telescoping gage. These gages must also be used in sets to cover a wide range of hole sizes but, unlike the telescoping gages, holes down to about 1/8 inch diameter can be measured. Sets of calibrated pins of known diameter may also be used to measure hole diameters. Pins of various sizes are inserted into the hole until a close, but not tight, fit is obtained. The hole diameter is then taken as the pin diameter. Pins are available in virtually any size, and are generally graduated in 0.0005 inch increments. Very extensive sets are needed to cover a range of nominal hole sizes. Of the devices available, calipers or small hole gages are most useful and economical for measuring hole diameters in composite test specimens. 6.4.2.8 Calibration of dimensional measurement devices In order to maintain the stated accuracy of mechanical measuring devices such as micrometers and calipers, they must be periodically calibrated. In general, there are no detailed calibration procedures available in high level (ASTM, ANSI, etc.) U.S. standards. Typically these instruments are calibrated using gage blocks, and specific procedures are contained in company internal specifications. Some ISO documents address aspects of this subject, and the reader is referred to standards under the jurisdiction of ISO Technical Committee 3 on Limits and Fits, as well as to ISO 10012-1 (Reference 6.4.2.8). 6.4.3 Load measurement devices 6.4.3.1 Introduction The ability to accurately and repeatably measure load (force) is critical to the testing and characterization of composite materials. This section will discuss the various types of instrumentation used to make load measurements, and provide guidelines to insure the accuracy of those measurements. Load meas-
MIL-HDBK-17-1F Volume 1,Chapter 6 Lamina,Laminate,and Special Form Characterization urement device classification and verification is discussed in ASTM E4"Standard Practices for Force Veri- fication of Testing Machines"(Reference 6.4.3.1(a)),ASTM E74 "Standard Practice of Calibration of Force-Measuring Instruments for Verifying the Force Indication of Testing Machines"(Reference 6.4.3.1(b)),and ASTM E467 "Standard Practice for Verification of Constant Amplitude Dynamic Loads on Displacements in an Axial Load Fatigue Testing System"(Reference 6.4.3.1(c)).Calibration of load de- vices is also discussed in ISO 5893"Rubber and plastics test equipment-Tensile,flexural and compres- sive types(constant rate of traverse)-Description"(Reference 6.4.3.1(d)). Note:Force in the case of testing machines is defined as pound-force,or Newton where one pound force is the force required to provide a one pound mass an acceleration of 32.1740 ft/sec2(9.80665 m/sec),and a Newton is the force required to provide a one kilogram mass an acceleration of 1m/sec. This force is used to determine the load applied on test specimens.Load is commonly used inter- changeably with force in mechanical testing specifications and in MIL-HDBK-17. 6.4.3.2 Load cells The most common type of force measurement device in the mechanical properties testing laboratory is the strain-gage instrumented load cell.These devices consist of an elastic member that deflects in a uniform,consistent,and repeatable manner under the application of load.The elastic member in the load cell is instrumented with strain gages so as to measure the deflection.The output of the strain gage cir- cuit can easily be read by a variety of recording devices and data acquisition systems.The strain gages in the load cell form a complete bridge,carefully balanced,so that the load cell can be calibrated using a reference excitation voltage,and thereafter the output of the bridge circuit will be dependent only upon the external conditioning circuitry.The bridge is,therefore,guaranteed to be in a balanced condition(at ther- mal equilibrium)when no load is applied.Load cells with internal signal conditioning circuitry should be avoided when circumstances may require the heating and/or cooling of the load cell.An important factor in the design or choice of a load cell is the ability of the load cell to reject spurious inputs generated by improper but inevitable misuse,such as off-axis loading and heating/cooling of the load cell during a test. The ability of a load cell to reject off-axis loads and thermal drift is dependent on the design of the elastic member and the placement of the strain gages upon that member.(See Section 6.4.2.4 on strain gages) A well designed load cell can have a repeatability of 0.01%(of the full-scale output of the load cell),and a thermal stability of 0.001%(full scale)per degree F. 6.4.3.2.1 Design and specification considerations Load cells should be chosen to provide the greatest degree of accuracy consistent with the required data.An indicated load accurate to within 0.1%of the actual load at critical points in the test(modulus chord points,failure load)will guarantee high quality test results.A variety of load cell configurations is available: 1.Bending beam load cells are constructed as a simple cantilever beam,with strain gages attached to measure deflection.The beam may be instrumented with a single strain gage(quarter bridge),two gages (half bridge),or four gages (full bridge).When two gages are used,they are connected in such a manner that the strains in the gages are summed,effectively doubling the sensitivity of the cir- cuit.When four gages are used,they may be arranged so as to quadruple the sensitivity,or to com- pensate for the nonlinearity of the strain gradient in the beam.Bending beam load cells are used when cost is a factor,as when destruction of the load cell is a possibility.High accuracy is possible with this type of load cell when it is used correctly.The"S"beam load cell is a special form of the bending beam load cell which permits "in-line"loading to be used with an inexpensive load cell de- sign.Bending beam load cells can reject torsional loading of the beam,and thermal effects,but at higher strains some designs become markedly nonlinear while still producing repeatable results. 2.Shear beam load cells,in their simplest form,utilize the uniform shear condition in the web of an I-beam shaped member as the surface of measurement.Precision load cells commonly utilize eight or twelve mechanical elements of the shear beam type arranged in a radially symmetric pattern, 6-13
MIL-HDBK-17-1F Volume 1, Chapter 6 Lamina, Laminate, and Special Form Characterization 6-13 urement device classification and verification is discussed in ASTM E4 “Standard Practices for Force Verification of Testing Machines” (Reference 6.4.3.1(a)), ASTM E74 “Standard Practice of Calibration of Force-Measuring Instruments for Verifying the Force Indication of Testing Machines” (Reference 6.4.3.1(b)), and ASTM E467 “Standard Practice for Verification of Constant Amplitude Dynamic Loads on Displacements in an Axial Load Fatigue Testing System” (Reference 6.4.3.1(c)). Calibration of load devices is also discussed in ISO 5893 “Rubber and plastics test equipment – Tensile, flexural and compressive types (constant rate of traverse) – Description” (Reference 6.4.3.1(d)). Note: Force in the case of testing machines is defined as pound-force, or Newton where one pound force is the force required to provide a one pound mass an acceleration of 32.1740 ft/sec2 (9.80665 m/sec2 ), and a Newton is the force required to provide a one kilogram mass an acceleration of 1m/sec. This force is used to determine the load applied on test specimens. Load is commonly used interchangeably with force in mechanical testing specifications and in MIL-HDBK-17. 6.4.3.2 Load cells The most common type of force measurement device in the mechanical properties testing laboratory is the strain-gage instrumented load cell. These devices consist of an elastic member that deflects in a uniform, consistent, and repeatable manner under the application of load. The elastic member in the load cell is instrumented with strain gages so as to measure the deflection. The output of the strain gage circuit can easily be read by a variety of recording devices and data acquisition systems. The strain gages in the load cell form a complete bridge, carefully balanced, so that the load cell can be calibrated using a reference excitation voltage, and thereafter the output of the bridge circuit will be dependent only upon the external conditioning circuitry. The bridge is, therefore, guaranteed to be in a balanced condition (at thermal equilibrium) when no load is applied. Load cells with internal signal conditioning circuitry should be avoided when circumstances may require the heating and/or cooling of the load cell. An important factor in the design or choice of a load cell is the ability of the load cell to reject spurious inputs generated by improper but inevitable misuse, such as off-axis loading and heating/cooling of the load cell during a test. The ability of a load cell to reject off-axis loads and thermal drift is dependent on the design of the elastic member and the placement of the strain gages upon that member. (See Section 6.4.2.4 on strain gages). A well designed load cell can have a repeatability of 0.01% (of the full-scale output of the load cell), and a thermal stability of 0.001% (full scale) per degree F. 6.4.3.2.1 Design and specification considerations Load cells should be chosen to provide the greatest degree of accuracy consistent with the required data. An indicated load accurate to within 0.1% of the actual load at critical points in the test (modulus chord points, failure load) will guarantee high quality test results. A variety of load cell configurations is available: 1. Bending beam load cells are constructed as a simple cantilever beam, with strain gages attached to measure deflection. The beam may be instrumented with a single strain gage (quarter bridge), two gages (half bridge), or four gages (full bridge). When two gages are used, they are connected in such a manner that the strains in the gages are summed, effectively doubling the sensitivity of the circuit. When four gages are used, they may be arranged so as to quadruple the sensitivity, or to compensate for the nonlinearity of the strain gradient in the beam. Bending beam load cells are used when cost is a factor, as when destruction of the load cell is a possibility. High accuracy is possible with this type of load cell when it is used correctly. The “S” beam load cell is a special form of the bending beam load cell which permits “in-line” loading to be used with an inexpensive load cell design. Bending beam load cells can reject torsional loading of the beam, and thermal effects, but at higher strains some designs become markedly nonlinear while still producing repeatable results. 2. Shear beam load cells, in their simplest form, utilize the uniform shear condition in the web of an I-beam shaped member as the surface of measurement. Precision load cells commonly utilize eight or twelve mechanical elements of the shear beam type arranged in a radially symmetric pattern
MIL-HDBK-17-1F Volume 1,Chapter 6 Lamina,Laminate,and Special Form Characterization which combined with a well designed bridge circuit utilizing four of the shear beams,allows the load cell to reject off-axis loads. 3.Ring load cells,essentially so-called proving rings,consist of an elastic member of a ring shape, which when loaded at diametrically opposite points deforms elliptically.This type of load cell can be of high accuracy,but does a poor job of rejecting off axis loads. 6.4.3.3 Other load measuring systems The following is a brief summary of other types of load measuring devices sometimes used.These systems are generally for highly specialized uses,or are based on older technology,and are not preferred for obtaining MIL-HDBK-17 data. LVDT Devices--A load cell which uses an LVDT (linear variable differential transformer,see 6.7.2.4.4)as a strain measuring device may occasionally be seen.This type of load cell may be as accurate as a bonded strain gage type cell,but is somewhat less rugged. Solid state load transducers--Special purpose load cells utilizing piezoelectric or piezoresistive semi- conductor strain measuring elements(see "Strain Gage Technology"(Reference 6.4.3.3)are available for measuring load during impact,when the strain change rate might exceed the ability of a bonded foil strain gage load cell to accurately indicate load.Semiconductor strain gages are extremely sensitive to tem- perature changes,and will yield rapid zero shifts with changing temperature.Therefore,they must be used only at thermal equilibrium. Bourdon tubes,etc.--There are older test machines in everyday use which rely on Bourdon tubes and other ingenious mechanisms for indicating load.A Bourdon tube is a sealed tube,formed in a spiral, semi-circle or helical shape and filled with fluid.When pressurized,the fluid causes the tube to move in a reproducible manner,mechanically acting on a readout device or indicating needle.The indicating dials on these machines should be relied on only as relative indicators of load level.In all cases,these ma- chines should be retro-fitted with electronic load cells and indicators which can be calibrated more readily. and to greater degrees of accuracy. Calibrated weights--Creep testing machines are commonly of the unequal arm dead-weight loading type.The weights used on these machines are commonly cast iron and should be calibrated to Class 6 of ASTM E 617(note this is the least precise class given in ASTM E 617 and has a tolerance of 0.01%). The measurement of the length of the arms and the condition of the knife-edges should be verified per the machine manufacturers instructions Levers --Test machines with an integral system of levers and knife edges for indicating load as on a beam balance or compound scale should be retrofitted with electronic load indicators to simplify calibra- tion and data acquisition. 6.4.3.4 Instrumentation and calibration Calibration(more properly,verification)of test machines and load cells requires a "CLASS A"load standard.Those standards are commonly high precision load cells or proving rings.The load standard must have an uncertainty not exceeding 0.25%of the load being measured.Therefore,the minimum load which it may be used to calibrate must be at least 400 times the uncertainty.ASTM E 4 allows the uncer- tainty of a load device used for testing to be up to 1%of full scale.ASTM E 74 contains a detailed expla- nation and analysis of the calibration of load measuring devices,and it should be studied closely by any- one responsible for calibrating these devices.Machines meeting the ISO requirements have a sliding scale of allowable error with a maximum of +1.0%at full scale for Grade A machines and +2.0%of full scale for grade B machines.A load cell calibrated to ISO 5893.Grade A meets the load cell requirements of ASTM E 4,though there are additional requirements in ASTM E 4 for test machines that are not cov- ered in the ISO standard.The ISO and ASTM standards both include other details such that they are not strictly interchangeable. 6-14
MIL-HDBK-17-1F Volume 1, Chapter 6 Lamina, Laminate, and Special Form Characterization 6-14 which combined with a well designed bridge circuit utilizing four of the shear beams, allows the load cell to reject off-axis loads. 3. Ring load cells, essentially so-called proving rings, consist of an elastic member of a ring shape, which when loaded at diametrically opposite points deforms elliptically. This type of load cell can be of high accuracy, but does a poor job of rejecting off axis loads. 6.4.3.3 Other load measuring systems The following is a brief summary of other types of load measuring devices sometimes used. These systems are generally for highly specialized uses, or are based on older technology, and are not preferred for obtaining MIL-HDBK-17 data. LVDT Devices -- A load cell which uses an LVDT (linear variable differential transformer, see 6.7.2.4.4) as a strain measuring device may occasionally be seen. This type of load cell may be as accurate as a bonded strain gage type cell, but is somewhat less rugged. Solid state load transducers -- Special purpose load cells utilizing piezoelectric or piezoresistive semiconductor strain measuring elements (see ”Strain Gage Technology” (Reference 6.4.3.3) are available for measuring load during impact, when the strain change rate might exceed the ability of a bonded foil strain gage load cell to accurately indicate load. Semiconductor strain gages are extremely sensitive to temperature changes, and will yield rapid zero shifts with changing temperature. Therefore, they must be used only at thermal equilibrium. Bourdon tubes, etc. -- There are older test machines in everyday use which rely on Bourdon tubes and other ingenious mechanisms for indicating load. A Bourdon tube is a sealed tube, formed in a spiral, semi-circle or helical shape and filled with fluid. When pressurized, the fluid causes the tube to move in a reproducible manner, mechanically acting on a readout device or indicating needle. The indicating dials on these machines should be relied on only as relative indicators of load level. In all cases, these machines should be retro-fitted with electronic load cells and indicators which can be calibrated more readily, and to greater degrees of accuracy. Calibrated weights -- Creep testing machines are commonly of the unequal arm dead-weight loading type. The weights used on these machines are commonly cast iron and should be calibrated to Class 6 of ASTM E 617 (note this is the least precise class given in ASTM E 617 and has a tolerance of 0.01%). The measurement of the length of the arms and the condition of the knife-edges should be verified per the machine manufacturers instructions. Levers -- Test machines with an integral system of levers and knife edges for indicating load as on a beam balance or compound scale should be retrofitted with electronic load indicators to simplify calibration and data acquisition. 6.4.3.4 Instrumentation and calibration Calibration (more properly, verification) of test machines and load cells requires a “CLASS A” load standard. Those standards are commonly high precision load cells or proving rings. The load standard must have an uncertainty not exceeding 0.25% of the load being measured. Therefore, the minimum load which it may be used to calibrate must be at least 400 times the uncertainty. ASTM E 4 allows the uncertainty of a load device used for testing to be up to 1% of full scale. ASTM E 74 contains a detailed explanation and analysis of the calibration of load measuring devices, and it should be studied closely by anyone responsible for calibrating these devices. Machines meeting the ISO requirements have a sliding scale of allowable error with a maximum of ±1.0% at full scale for Grade A machines and ±2.0% of full scale for grade B machines. A load cell calibrated to ISO 5893, Grade A meets the load cell requirements of ASTM E 4, though there are additional requirements in ASTM E 4 for test machines that are not covered in the ISO standard. The ISO and ASTM standards both include other details such that they are not strictly interchangeable
MIL-HDBK-17-1F Volume 1,Chapter 6 Lamina,Laminate,and Special Form Characterization 6.4.3.5 Precautions Certain precautions should be observed to insure the accuracy of load cell readings. a.In all cases,scrutinize the specifications of a commercial load cell,or carefully analyze the bridge cir- cuit of a self manufactured load cell.Curve fitting of nonlinear output is possible,but care should be taken to insure that the fitting equation is correct,and that it is applied correctly. b.The load cell should be calibrated at regular intervals to verify its performance per ASTM E 4 and E 74.The calibration device should be traceable to a national standard such as NIST,and its readout and accuracy should exceed that of the device being calibrated by a factor of 4 or more. c.The load cell calibration should be reverified whenever unusual loading conditions occur,such as overloading,impacting,or bending (off-axis loading).Refer to the specification of the individual load cell for overload tolerances. d.Load cells not specifically declared to be tolerant of temperature changes during testing should be assumed to be inaccurate at elevated or depressed temperatures.Therefore,care should be taken to isolate the load cell and its cabling,from temperature changes and/or gradients. e. Care must be taken to insure that the applied load axis corresponds as nearly as possible to the indi- cated loading axis of the load cell.Off-axis loads should be avoided and may result in inaccurate readings,and may damage the load cell or other parts of the load train. f.In general,the capacity of the load cell to be used for a given test should be determined such that the predicted failure load is between 15%and 85%of the capacity of the load cell.If expected loads are less than 15%of the load cell capacity,the user should insure that adequate calibration has been per- formed in the test range.Such a calibration may be outside the scope of a routine ASTM E 4 calibra- tion,so special arrangements may need to be made.When the instrumentation used permits the 'ranging"of the load cell,for instance where a 100,000 pound load cell might be used over a range of 10,000 pounds through amplification,the load cell must be separately calibrated under those circum- stances as a 10,000 pound capacity load cell.Similar individual calibrations must be conducted for all "ranges"provided.Use of a load cell for testing when the expected data is greater than 85%of the load cell capacity is discouraged since an unexpected high load may exceed the capacity of the load cell. 6.4.4 Strain/displacement measurement devices 6.4.4.1 Introduction The ability to accurately and repeatability measure deformation and displacement is critical to the testing and characterization of composite materials.This section will discuss the various types of instru- mentation used to make strain measurements,and provide guidelines to help determine the appropriate methods for various test types,material forms,test conditions,and data requirements.Extensometer classification and verification is discussed in ASTM E 83(Reference 6.4.1).The class of the extensom- eter is determined from the maximum expected error.Class A has the least expected error,followed by classes B-1,B-2,C,D,and E in that order.Calibration to class A is very difficult to achieve.Only those extensometers which can be classified as ASTM E 83 Class B-2 or better are acceptable for generating data to be included in MIL-HDBK-17. 6-15
MIL-HDBK-17-1F Volume 1, Chapter 6 Lamina, Laminate, and Special Form Characterization 6-15 6.4.3.5 Precautions Certain precautions should be observed to insure the accuracy of load cell readings. a. In all cases, scrutinize the specifications of a commercial load cell, or carefully analyze the bridge circuit of a self manufactured load cell. Curve fitting of nonlinear output is possible, but care should be taken to insure that the fitting equation is correct, and that it is applied correctly. b. The load cell should be calibrated at regular intervals to verify its performance per ASTM E 4 and E 74. The calibration device should be traceable to a national standard such as NIST, and its readout and accuracy should exceed that of the device being calibrated by a factor of 4 or more. c. The load cell calibration should be reverified whenever unusual loading conditions occur, such as overloading, impacting, or bending (off-axis loading). Refer to the specification of the individual load cell for overload tolerances. d. Load cells not specifically declared to be tolerant of temperature changes during testing should be assumed to be inaccurate at elevated or depressed temperatures. Therefore, care should be taken to isolate the load cell and its cabling, from temperature changes and/or gradients. e. Care must be taken to insure that the applied load axis corresponds as nearly as possible to the indicated loading axis of the load cell. Off-axis loads should be avoided and may result in inaccurate readings, and may damage the load cell or other parts of the load train. f. In general, the capacity of the load cell to be used for a given test should be determined such that the predicted failure load is between 15% and 85% of the capacity of the load cell. If expected loads are less than 15% of the load cell capacity, the user should insure that adequate calibration has been performed in the test range. Such a calibration may be outside the scope of a routine ASTM E 4 calibration, so special arrangements may need to be made. When the instrumentation used permits the “ranging” of the load cell, for instance where a 100,000 pound load cell might be used over a range of 10,000 pounds through amplification, the load cell must be separately calibrated under those circumstances as a 10,000 pound capacity load cell. Similar individual calibrations must be conducted for all “ranges” provided. Use of a load cell for testing when the expected data is greater than 85% of the load cell capacity is discouraged since an unexpected high load may exceed the capacity of the load cell. 6.4.4 Strain/displacement measurement devices 6.4.4.1 Introduction The ability to accurately and repeatability measure deformation and displacement is critical to the testing and characterization of composite materials. This section will discuss the various types of instrumentation used to make strain measurements, and provide guidelines to help determine the appropriate methods for various test types, material forms, test conditions, and data requirements. Extensometer classification and verification is discussed in ASTM E 83 (Reference 6.4.1). The class of the extensometer is determined from the maximum expected error. Class A has the least expected error, followed by classes B-1, B-2, C, D, and E in that order. Calibration to class A is very difficult to achieve. Only those extensometers which can be classified as ASTM E 83 Class B-2 or better are acceptable for generating data to be included in MIL-HDBK-17