Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
CONTENTS SECTION 3 LABORATORY TESTING...............................................................................................................3-1 3.1 INTRODUCTION ........................................................................................................................................3-2 3.2 TESTS OF WIRE PROPERTIES...............................................................................................................3-2 3.2.1 Specimen Preparation ........................................................................................................................3-2 3.2.2 Tensile Tests ........................................................................................................................................3-3 3.2.3 Obtaining Data for Force vs. Strain Curves.....................................................................................3-3 3.2.4 Fractographic Examination of Suspect Wires .................................................................................3-3 3.2.5 Examination of Fracture Surface for Preexisting Cracks...............................................................3-4 3.3 ZINC COATING TESTS.............................................................................................................................3-5 3.3.1 Weight of Zinc Test ............................................................................................................................3-5 3.3.2 Preece Test for Uniformity.................................................................................................................3-5 3.4 CHEMICAL ANALYSIS.............................................................................................................................3-6 3.5 CORROSION ANALYSIS ..........................................................................................................................3-6 3.6 FIGURES FOR SECTION 3 .......................................................................................................................3-8 3.7 REFERENCE ................................................................................................................................................3-8 3-1
GUIDELINES COMMENTARY 3-2 3.1 INTRODUCTION Laboratory testing is an integral part of cable inspection. The test results are used to estimate the strength of the wires and their stress vs. strain relationships, which are used in turn to evaluate cable strength. The same tests are used to determine the ultimate strain of the wires for the Limited Ductility Model. Other tests assess the remaining life of the zinc coating. Additional tests are performed on cable wires to study the causes of corrosion. Although they are referred to in this section, they are irrelevant to the assessment of structural safety. 3.2 TESTS OF WIRE PROPERTIES Strength testing is the most essential type of testing for the evaluation of cable capacity. 3.2.1 Specimen Preparation A sample wire is a length of wire that has been removed from a cable for testing. A specimen is a piece of wire cut from the sample on which a specific test is performed. Sample wires obtained in the field should be long enough to provide the number of specimens recommended in Table 2.4.3.5.3-1. All of the specimens from a given sample should be at the same stage of corrosion, but it is understood this is not always possible. The cast diameter should be determined prior to cutting specimens from the sample wires. If the sample is of sufficient length to form a complete circle as it lies on a flat surface, measure the cast diameter in two perpendicular directions and average the results. If the sample is not long enough to form a complete circle, measure the rise of the arc on each of two convenient chords of the curve, calculate the resulting diameters geometrically, and average the results. The diameter d is given by b cbd 8 42 22 + â = 3.2.1-1 where b = offset between chord and arc c = chord length C3.2.1 A typical Stage 4 sample with no cracks has a standard deviation that is approximately 1% to 2% of the mean tensile strength. Ten specimens are sufficient to determine the sample mean tensile strength within 3% of the true mean with a 97.5% confidence level. This number of specimens cannot be obtained during the first internal cable inspection, if the recommended 16 feet of cable are unwrapped. Longer lengths of cable should be unwrapped whenever corrosion is found to exceed Stage 3, so that Stage 4 wire samples that are at least 16 feet long can be removed. Whenever corrosion is found not to exceed Stage 3, cracks are not likely to be present, and 12-foot-long samples providing 8 Stage 3 specimens are adequate. Wires must be cleaned and all corrosion product must be removed prior to using a dye penetrant to find cracks. Even then, shallow pitting may obscure cracks or be confused with them. The most reliable method of identifying cracks is to inspect the fracture surface visually after testing. Optical microscopy prior to testing is the alternative to using dye or magnetic flux leakage. It is extremely time-consuming, and the human eye can fail to spot many cracks, even with a microscope.
GUIDELINES COMMENTARY 3-3 Before sample wires are cut into specimens of suitable length for testing, they should be inspected and assigned to the appropriate corrosion stage. If possible, NDE testing to locate preexisting cracks should be performed on individual wires before they are cut, so that the worst cracks can be arranged to appear near the center of the specimen. Among the techniques that may be used to identify cracks are the application of a dye penetrant on cleaned wires and magnetic flux leakage inspection. 3.2.2 Tensile Tests Wire strength derived from tensile tests is used to estimate cable strength. Tensile tests should be performed in accordance with ASTM A586 and ASTM A370 to determine the following wire properties: breaking load in the wire yield strength (0.2% offset method) tensile strength elongation in 10-inch-gage length reduction of area modulus of elasticity The tensile strength should be based on the nominal area of the wire. C3.2.2 Many engineers prefer to use wire tensile strength rather than wire strength (a force) in calculations, and therefore strength equations in the Guidelines are derived using tensile strength, which is multiplied by the nominal area of the wire to calculate cable strength. For this reason, tensile strength test results should be based on the nominal area as well. Either gross metallic area (including the area of the zinc coating) or net steel area (not including the area of the zinc coating, which is equivalent to the nominal area of the uncoated wire) may be used, as long as the same area is used consistently in all cable strength calculations. The zinc coating is often degraded in the samples removed from the cable and the net steel area is preferred for the calculations, because the actual diameter of the wire without galvanizing can be measured more accurately in the laboratory. Whenever section loss is observed in the specimen, the stress in the actual corroded wire cross-section may be of interest, because it provides the actual tensile strength of the steel in the corroded area. Tests on wires from the anchorages of the Manhattan Bridge have shown that this value does not change when section loss occurs. Significant variation may indicate hydrogen embrittlement cracking or pitting of the wires. 3.2.3 Obtaining Data for Force vs. Strain Curves In addition to the tests listed above, wire elongation should be recorded at intervals of tensile force up to maximum force preceding failure. The data should be used to construct a full stress-strain curve, or force vs. strain curve, for each specimen. The ultimate strain corresponding to tensile strength should be determined as well. C.3.2.3 The testing laboratory selects the technique for determining elongations beyond 2%, because extensometers can be damaged whenever the wire fails. One option is to measure the motion of the head of the testing machine, adjusting the elongation for slippage at the time the grips are set. The measurement should include both the elastic and plastic components of the deformation. 3.2.4 Fractographic Examination of Suspect Wires The fracture surface of the wires should be observed C3.2.4 Techniques that are generally used to study the microstructure of metal can also be used to study failure
GUIDELINES COMMENTARY 3-4 to detect whether failure is ductile or brittle. A brittle failure is consistent with pitting or cracking, loss of ductility, a reduction in elongation and strength, and little or no reduction in area. Special attention should be focused on the causes of these phenomena. The instruments recommended for the task are a stereoscopic optical microscope and/or a scanning electron microscope. It is also recommended that X-ray energy dispersion spectral analysis be performed on any fracture surface that displays traces of corrosion or contamination. and corrosion morphologies. ⢠Optical (Light) Microscopy A stereoscopic microscope with 20X magnification is the most efficient tool for the detection of preexisting cracks in the fracture surface. Crack depth can be measured directly, if the microscope is fitted with a reticle, or indirectly by taking a microphotograph of the fracture surface. Longitudinal sections of wire that are microetched may be studied with an optical microscope at magnifications of 50X to 200X to identify corrosion morphologies in pits, both intergranular and intragranular, and to establish the paths of secondary cracks near the fracture surface. ⢠Scanning Electron Microscopy (SEM) Any characteristic that can be studied with optical microscopy is more easily studied with SEM. SEM allows greater depth of field and better resolution at higher enlargements than optical microscopy. At the same time, SEM allows simultaneous close study of the failed surface. With SEM, failure surfaces are visible with sufficient detail to identify the failure mechanism, such as cleavage or ductile rupture. Corrosion residues are also discernible on the metal structure or fracture surface. ⢠Machining, Polishing and Etching Bridge wires should be machined and polished in the plane of the cast radius. This method generally cuts across any transverse pits and cracks that initiate at the inner radius. The polished surface is etched with various reagents to reveal the steel microstructure in detail. ⢠Image Interpretation Images of failure morphologies under enlargement should be interpreted by metallurgists or, if they are unusual, by corrosion experts. The images may indicate embrittlement, hydrogen-assisted cracking or other corrosion mechanisms, recognizable to experts in these fields. 3.2.5 Examination of Fracture Surface for Preexisting Cracks Cracked wires are treated as a separate group in estimating cable strength. Preexisting cracks are defined as cracks that are present in the specimen prior to testing. They are found by examining the fracture surface of all tension specimens under a C3.2.5 In Stage 1 and Stage 2 wires, preexisting cracks are usually due to a manufacturing flaw, while in Stage 3 and Stage 4 wires, the cracks are most often caused by hydrogen that results from galvanic action. The surfaces of the preexisting cracks in the Stage 3 and Stage 4 wires are usually black, and the cracks themselves are easily
GUIDELINES COMMENTARY 3-5 stereoscopic optical microscope at 20X magnification. A sample wire is considered to contain a crack if any of the specimens cut from the sample contains a preexisting crack. The outer surfaces of the wire in the vicinity of a brittle fracture should also be examined under a stereoscopic optical microscope for the presence of additional preexisting cracks. A cracked specimen should be photographed, and the crack depth should be measured. The wire diameter at the failure plane, as well as the crack depth, should be reported in both absolute terms and as a fraction of wire diameter. Longitudinal sections of short wire segments in the vicinity of a brittle fracture should be examined under either an optical or scanning electron microscope. In preparation, the surface of the specimen section should be polished and etched. The recommended etchant is a 10% solution of nitric acid in ethyl alcohol. distinguished from the fracture surfaces caused by the testing load. A cracked wire is shown in Figure 3.2.5-1. 3.3 ZINC COATING TESTS Two types of tests are performed on the zinc coating during cable wire evaluation: Weight of Zinc Coating Tests and Preece Tests. The minimum depth of the coating determines its condition, not the average depth. C.3.3 Wires often display white spots on a shiny silvery field of sound zinc. If the white spots represent 30% of the surface area or more, then there may be significant variations in the depth of the zinc coating. 3.3.1 Weight of Zinc Test The Weight of Zinc Coating Test, specified in ASTM A90, is a gravimetric test that measures the weight of the zinc removed from a unit length of wire. It is used to determine the average weight of zinc in that length, separate from variations in coating thickness. Weight of Zinc Coating Tests should be conducted on Stage 1 and Stage 2 specimens that display uniform zinc or spotted zinc loss. C3.3.1 The average weight of zinc in a unit length, determined by testing, can be converted to an average remaining thickness of zinc coating and used to predict when the zinc coating will be depleted. 3.3.2 Preece Test for Uniformity The Preece Test, specified in ASTM A239, is used to determine the uniformity of the zinc coating on Stage 1 and Stage 2 wires. Preece Tests are chemical tests that depend on the reaction of copper sulfate and zinc. They are used to confirm whether the coating on the specimen is depleted uniformly or locally. Preece Tests should be conducted on Stage 1 and C3.3.2 Preece Tests are performed in series. Wires are dipped in a copper sulfate solution for a standard time period. If sufficient zinc is present, then the wire retains its shiny surface from the intact zinc. If the zinc is insufficient, then the copper electroplates the steel, and the wire surface turns the color of copper. The tests are terminated after the fourth dip.
GUIDELINES COMMENTARY 3-6 Stage 2 specimens that display uniform zinc or spotted zinc loss. 3.4 CHEMICAL ANALYSIS The chemical composition of the steel wire should be determined under any of these circumstances: tests were never performed, results from previous tests are unavailable, or tests reveal unusual variations in the tensile strength of samples. Percentages of the following elements should be obtained: ⢠carbon ⢠silicon ⢠manganese ⢠phosphorous ⢠sulfur ⢠copper ⢠nickel ⢠chromium ⢠aluminum Five wires should be analyzed to provide a complete record for future inspections. If the chemistry of the steel is found to vary significantly, a metallurgist should be consulted to study the effects on the properties of the wire. A chemical analysis of the surface deposits on the wire samples should be performed if corrosion is present, to detect harmful contaminants. The presence or absence of the following salts should be established: ⢠chloride ⢠sulfates ⢠nitrates The results should be reported in absolute amounts, per unit of wire area. C3.4 Variations in the carbon content of the wires may cause wider than usual variations in tensile strength. The Williamsburg Bridge cable wires are an example of this phenomenon. In ensuing inspections of such bridges, the pattern of sample taking should differ from the recommended pattern so that the extent of the variation in carbon content, the tensile strength and the ultimate strain throughout the cable can be determined. The proper procedure to follow is described by Matteo [1]. Aluminum, the last element listed, is not usually present in bridge wire, unless it has been used as a killing agent in the production of the steel. 3.5 CORROSION ANALYSIS In some cases, the investigator may recommend studying the corrosion product on a wire or anchorage. Corrosion analysis can be performed on surface corrosion films, or on the fracture surfaces of the steel, or on corrosion by-products. C3.5 Various types of electronic microscopy are used in corrosion analysis: ⢠X-Ray Photoelectron Spectroscopy (ESCA) Also referred to as Electron Spectroscopy for Chemical Analysis or ESCA, X-Ray
GUIDELINES COMMENTARY 3-7 Chlorides from roadway salts; sulfates, and nitrates from acid rain are some of the causes of corrosion revealed by the analysis. Remedial measures may be recommended to eliminate polluting elements. Photoelectron Spectroscopy is a surface-sensitive spectroscopic technique that provides information about the composition and structure of the outermost atomic layers (2 nm) of a solid material. ESCA detects all elements except hydrogen and helium. The element detection limit is typically about 0.5%. Sometimes it is possible to determine the chemical state of elements, including their bonding structure, using this technique. ⢠Energy Dispersive X-Ray Analysis (EDAX) The Energy Dispersive X-Ray Spectrometer is an attachment to the Scanning Electron Microscope that identifies elements on the surface from X- rays emitted by the specimen. EDAX can detect elements as light as boron (atomic number 5). It is particularly suited to identifying inorganic elements. The results are only semi-quantitative without the use of primary standards, which are recommended. This is due to the complex combinations of variables, such as sample size, surface condition, and orientation of the apparatus. However, the small peak-to- background ratio encountered in analysis of low concentrations of elements is an unavoidable occurrence that makes adequate quantitative analysis nearly unobtainable. ⢠X-Ray Diffraction (XRD) X-Ray Diffraction is used to obtain information about the structure, composition and state of polycrystalline materials. It can be used to determine the exact composition and state of the corrosion products. For instance, if adequate amounts of the product are available, it can identify various oxides of a particular element (e.g., magnetite Fe3O4 and hematite Fe2O3).
3.7 REFERENCE 1. Matteo, J., G. Deodatis, and D. P. Billington, Safety Analysis of Suspension-Bridge Cables: Williamsburg Bridge. Journal of Structural Engineering, 1994. 120(11): p. 3197â3211. 3-8 CRACK Figure 3.2.5-1. Cracked wire. 3.6 FIGURE FOR SECTION 3
