Nội dung

EM 1110-2-6054 1 Dec 01 (3) In 1964, a new specification, ASTM A502, was published for steel structural rivets, and superseded ASTM A141 and A195. The later version of this specification (ASTM A502) covers three grades of steel rivets: general-purpose carbon steel rivets, carbon-manganese steel rivets for use with high-strength carbon and high-strength low-alloy steels, and rivets comparable to ASTM A588 weathering steel. The later specification includes hardness requirements but not tensile and yield strength requirements. c. Allowable and yield stresses. During the same period that A7 steel was evolving, the American Institute of Steel Construction (AISC) changed their basic allowable working stress for structural steel only once, raising it in 1936 from 124 to 138 MPa (18 to 20 ksi) (Ferris 1953). The ASTM requirement for minimum yield point during this period was generally one-half times the tensile strength, or not less than 207 MPa (30 ksi); in 1933, the minimum of 207 MPa (30 ksi) was raised to 227.5 MPa (33 ksi) for plate and shape products. When A373 steel was introduced, that steel had a minimum yield point of 220.6 MPa (32 ksi), suggesting that to improve weldability at that time, some sacrifice in strength was necessary. Only when A36 steel was introduced in 1960 in a tentative specification (ASTM A36-60T) did the minimum yield point for structural steel plates and shapes increase to 248 MPa (36 ksi). By that time, weldability and welding practices for structural steel had markedly improved and standardized. d. Weldability of earlier steels. (1) A very good reference that discusses the weldability of steels, including steels that have limited weldability, is the monograph “Weldability of Steels” published by the Welding Research Council (Stout et al. 1987). Now in its fourth edition, the monograph has chapters on the properties of steel related to weldability, factors affecting weldability in fabrication, and the weldability of different steels. (2) For early steels, reference can be made to the first edition of the monograph (Stout and Doty 1953) which includes suggested (as of 1953) welding practices for A7 steel meeting the tentative specification ASTM A7-50T. However, even the first edition does not include data for A9 or A94 steels. A copy of the suggested (1953) practices for A7 steel is listed in Table 7-1. For thicknesses up to 1 in. (the normal case for hydraulic steel structures), a comparison of the recommended practices in Table 7-1 suggests that for carbon levels of 0.25 percent or less, no special welding requirements are needed for A7 steel. However, as the carbon level increases, more stringent practices are needed. Because A7 steel did not have a specified carbon level, repair and maintenance welding should be conducted favoring the more stringent practices. For other early steels or for steels of unknown specification, ANSI/AWS D1.1 provides optional methods for determining welding requirements based on the chemical composition of the steel. (3) A generally conservative practice for repair and maintenance welding on riveted spillway gates is to use the practices for A7 steel in Table 7-1, with the assumption that the carbon level is between 0.26 and 0.30 percent. 7-23 EM 1110-2-6054 1 Dec 01 Table 7-1 Suggested Practices for Sound Welding with A7 Steel (After Stout and Doty 1953) 7-24 EM 1110-2-6054 1 Dec 01 Chapter 8 Repair Considerations 8-1. General Most damage to hydraulic structures is due to impact of barges and debris, corrosion, or cracking. Many hydraulic structures are riveted and may include damaged or loose rivets that must be replaced. Repairs to hydraulic steel structures must maintain the required structural integrity and should be designed, if possible, to avoid recurrence of the original damage. In all cases, repairs should be designed using industry-approved detailing and fabrication procedures and should be detailed to avoid future corrosion or cracking problems (see paragraph 8-2a and paragraph 8-3). Repair of corroded areas is discussed in paragraph 8-2b, repair of cracks is discussed in paragraph 8-4, and rivet replacement is discussed in paragraph 8-5. Paragraph 8-6 discusses several repair examples. The type of repair details selected will be determined considering the following factors: a. Cause of damage. If the cause of original damage is not accounted for, it is likely that the damage will reoccur. If possible, the cause of damage should be eliminated or minimized. b. Remaining service life of the structure. A repair of a structure that is intended to be in service for only a short time might obviously be less extensive than for a structure intended for longer service. c. Frequency and type of future inspections. Due to cost or construction constraints, it may be prohibitive to provide an ideal repair. In such cases, a less than ideal repair might be adequate provided a strict inspection plan is developed. Development of inspection schedules for fatigue cracking is discussed in paragraph 6-11. d. Construction constraints. In general, repairs must be completed in a field environment that will include less than ideal conditions. For example, access to the repaired area may be restricted. Conditions may not be appropriate for welding (i.e., temperature, water, or access may inhibit proper welding). Certain situations might involve decrease in structural strength resulting from temporary removal of rivets, cover plates, or other parts. Distribution of dead and live load stresses must be considered. Repair components are generally effective only in resisting live load unless dead load is removed during repair. Each of these conditions will influence the design of the repair detail. 8-2. Corrosion Considerations a. New repair details. The primary means to avoid corrosion is by providing a protective coating system. The coating system applied to repair plates or components must be compatible with the protective system of adjacent steel. EM 1110-2-3400, CWGS 09940 and CWGS 05036 provide detailed information on selection, application, and specifications of coating systems. Structural detailing also has a significant impact on susceptibility to corrosion. Repairs should be detailed as much as possible to compensate for conditions that contribute to corrosion (paragraph 3-3b). The following items should be considered in the design process: (1) Detail components such that all exposed portions of the repair detail can be painted properly. Break sharp corners or edges to allow paint to adhere properly. (2) Where repair plates or components are horizontal, provide drain holes to prevent entrapment of water. Drain holes should be located at the lowest position with the size generally ranging from 25 mm (1 in.) to 75 mm (3 in.) in diameter. The cut edges of holes should be smooth and free of notches especially in areas subject to tensile forces. 8-1 EM 1110-2-6054 1 Dec 01 (3) Grind weld ends, slag, weld splatter, or any other deposits off the steel. These are areas that form crevices that can trap water. Use continuous welds. (4) Where dissimilar metals are in contact (generally carbon steel and either stainless steel or bronze), provide an electric insulator between the two metals and avoid large ratios of cathode (stainless steel) to anode (carbon steel) area. Surfaces of both metals should be painted. (5) Welded connections are generally more resistant to corrosion than bolted connections. In bolted connections, small volumes of water can be trapped under fasteners and between plies that are not sealed. Where possible, use welds in lieu of bolts considering the effect on fracture resistance. b. Existing corroded components. Where significant corrosion exists but strengthening is not required, the area should be cleaned appropriately and a new protective coating system applied. This will inhibit further corrosion, and future repairs might be avoided. In many cases, gate components such as skin plates include pitting corrosion that reduces component thickness where pitting exists. In certain cases, pits may be repaired by filling with weld metal. If this is done, strict weld procedures must be specified so the process is compatible with the existing base metal. This method is not recommended for fracture-critical components. 8-3. Detailing to Avoid Fracture a. General considerations. Regarding fracture resistance and fatigue strength, bolted repairs are often preferable to welded repairs. However, bolted repairs typically are more expensive and require more time to design and install. Dimensional constraints can also restrict the use of bolted splice plates. Sound welding, particularly under field conditions encountered during repair operations, can be difficult, increasing the possibility of poor quality welds. Moisture, paint, and other foreign material that can produce weld defects and cracking are often present. Welding residual stresses and degradation of material toughness in the heat-affected zone can also contribute to cracking of the repair. Weld intersections, intermittent welds, and tack welds on tension members should be avoided. b. Distortion. Most of the fatigue damage detected in U.S. bridges is due to distortion, mainly at unstiffened web gaps at the ends of diaphragms or floor beam connection plates (Keating 1994: Fisher 1984). An excellent summary of case studies on bridge failures due to distortion is presented by Fisher (1984). Out-of-plane behavior has been measured in field testing of lift gates (Commander et al. 1994). Unintended distortion can result from unanticipated forces such as those occurring at a semirigid connection designed to be a simple connection. Unintended distortion is generally due to out-of-plane displacement of structural components that is not accounted for in design. Details that are known to be predisposed to distortion damage should be avoided. c. Better details. (1) Most fractures of structural elements are attributed to adverse stress concentration conditions, unintended distortion, and inferior fabrication. Regardless of the primary contributing factor, cracking is generally associated with or assisted by conditions of adverse stress concentration. Therefore, detailing to minimize the effect of stress concentrations will prevent most fractures or at least will provide a more durable condition. If fracture is a concern, the detail that provides the least critical stress concentration condition should be used. In the design of structural details for new or repair applications, utilization of fatigue design criteria is a very simple means to ensure high fracture resistance. Even without fatigue loading, susceptibility to fracture is reflected by the level of stress and the stress concentration condition as discussed in paragraph 3-3a. 8-2 EM 1110-2-6054 1 Dec 01 (2) Fatigue strength relationships (Sr-N curves) for welded details are described in paragraph 2-3b. All bolted details provide a lower bound strength equivalent to a Category B detail. Due to a lower clamping force in rivets, riveted details have lower fatigue strength compatible with Category C or Category D as described in paragraph 2-3c. (3) A designer has some flexibility in selecting a detail to minimize likelihood of cracking. For a given condition, various details would serve the same purpose but have different fatigue strength. If budget and other constraints permit, a designer should generally choose the detail with the highest fatigue strength (the least likely to have cracking problems) regardless of loading. In cases of repair, the goal should be to provide a condition with an improved fatigue resistance compared to the original condition. Figure 8-1 shows several situations where a detail can be improved. (4) If possible, all Category E details should be avoided. Figures 8-1a and 8-1b demonstrate that other details with higher fatigue strength may be substituted for Category E details. The gusset plate attachments on the left sketch in each figure produce a Category E in the girder flange at the termination of the longitudinal welds or where there is a transverse weld. The Category E situation can be avoided by using a bolted connection (category B) or a gusset plate with a full-penetration weld ground smooth and a generous radius. In both cases, the stress concentration condition has been improved dramatically. Another way to avoid the adverse effects of Category E details is to locate the detail in a region of low stress. Cover plates can be extended to a region of low flexure, and attachments to flanges can be moved to the web where the flexural stress is low. (5) Fatigue strength can be improved significantly by providing a smooth transition between connected elements as shown in Figures 8-1b and 8-1c. A flange attachment can be improved from a Category E to a Category B detail by providing a radius transition and grinding the weld smooth (Figure 8-1b). The fatigue strength of a transverse groove weld is improved from a Category C to a Category B by removing the weld reinforcement and grinding the weld smooth (Figure 8-1c) (see groove welded connections of Table 2-1). Other important considerations are to avoid intermittent welds on backup bars and discontinuous backup bars. A category E situation exists at the termination of each intermittent weld, and a built-in crack exists where a backup bar is discontinuous. 8-4. Repair of Cracks Effective methods for repairing cracks or improving a detail include weld-toe grinding, peening, remelting, and hole drilling. The appropriate repair method for a given situation is dependent on the size and location of the crack and the type of detail at which the cracking occurred. Small through-thickness cracks subject to low stress range can be arrested by drilling a hole at the crack tip. For large cracks and/or higher stress range, repair can be accomplished by removing the crack tip by drilling a hole and repairing the remaining length of the crack by welding or with bolted splice plates. Simply welding the crack closed, even with a full penetration weld, should never be done without removing the crack tip with a hole. Such a repair generally worsens the condition due to the added residual stresses and deleterious thermal effects of welding (see paragraph 2-2c). Shallow surface cracks that typically occur at the toe of fillet welds can be repaired by grinding, air hammer peening, or gas tungsten arc (GTA) remelting. Surface cracks with depths that exceed the penetration capability of GTA remelting and the effectiveness of peening cannot be repaired by these procedures. Such cracks can be repaired by installing bolted splice plates that transfer the stress around the crack. a. Hole drilling. Hole drilling is the most commonly applied means of arresting fatigue cracks. A hole drilled at the tip of a crack essentially blunts the crack tip and the local stress concentration is greatly reduced compared to that for a sharp crack. It has been successfully applied to various types of structures, including navigation lock gates and several bridges (Fisher 1984). Hole drilling is effective for through-thickness cracks in plates or plate components of structural members. 8-3 EM 1110-2-6054 1 Dec 01 Figure 8-1. Better fatigue details (1 in. = 2.54 cm) (after Fisher 1977) where tw is the web thickness, M is the bending moment, L is the length of weld considered, and R is the radius on attached component (Continued) 8-4 EM 1110-2-6054 1 Dec 01 Figure 8-1. (Concluded) (1) The minimum hole size required to prevent crack initiation can be estimated with the relationship ∆K r = 26.7 σ y (8-1) 8-5 EM 1110-2-6054 1 Dec 01 where ∆K = stress intensity factor range, MPa- m r = radius of the hole, m σy = yield stress, MPa (For non-SI units, ∆K r = 10 σ y where r is in in., σy is in ksi, and ∆K is in ksi- in. ) (2) ∆K is calculated considering the entire stress range (algebraic sum of the tensile and compressive stress), and the crack size a (see Equation 6-1) includes the extent of the hole. Equation 8-1 is valid for structural steel and provides reasonable results for moderate stress ranges (less than 40 MPa (6 ksi)) and crack sizes. For most practical cases regarding moderate crack size and stress range less than 40 MPa (6 ksi), a crack will not reinitiate from a hole if the hole diameter is at least one-fifth the total crack length (Keating 1994). Hole diameters of 20 mm (13/16 in.) and 27 mm (1-1/16 in.) are practical since these sizes are commonly used for installation of high-strength bolts. The following are appropriate steps to arrest a small crack in structural steel subject to moderate stress range: (a) Determine the appropriate hole size (Equation 8-1). (b) Locate the crack tip with dye penetrant testing. (c) Drill hole with center at crack tip. (d) Inspect drilled surface of hole with dye penetrant testing to verify complete removal of crack tip. (e) In some cases, crack reinitiation may be inhibited by installing tightened bolts in the holes. This introduces local compressive stresses in the through-thickness direction that inhibits crack initiation from the hole. (3) For larger crack sizes and stress range greater than 40 MPa (6 ksi) (often the case for hydraulic structures), the hole size required by Equation 8-1 is significant and is not practical. In these cases, the crack tip may be removed by drilling a hole and the remaining crack repaired by welding or bolted splice plates. The following are general guidelines for a welded crack repair: (a) Clean area and determine extent of crack with dye penetrant testing (see paragraph 4-5b). (b) Drill hole at crack tip location. (c) Gouge out crack and prepare joint as a full-penetration groove weld in accordance with ANSI/AWS D1.1. (d) Preheat and weld joint using runout tabs and backing as required per American National Standards Institute/American Welding Society (ANSI/AWS) D1.1. 8-6 EM 1110-2-6054 1 Dec 01 (e) Remove backing and runout tabs. (f) Grind weld smooth. (g) Ream hole to remove weld metal and smooth edges. (h) Verify removal of crack tip with dye penetrant testing. (i) Inspect weld with appropriate nondestructive testing (ultrasonic testing or radiographic testing (paragraph 4-5)). (4) Alternatively, a bolted repair of the fatigue crack can be installed after a hole is drilled to arrest the crack: (a) Determine extent of crack with dye penetrant testing. (b) Drill hole at crack tip location. (c) Verify removal of crack tip with dye penetrant testing. (d) Prepare and install bolted repair over the crack. b. Weld toe grinding. Weld-toe grinding reduces the geometrical stress concentration and extends the fatigue life of undamaged details (Keating 1994). Grinding can be used to remove shallow fatigue cracks that may exist in the weld toe. Grinding should always be done in the direction of applied stress. A pencil or rotary burr grinder can be used. Magnetic particle inspection of the ground area should be conducted after grinding to ensure that embedded flaws are not exposed. (Penetrant inspection may reveal false indications due to grinding marks.) c. Peening. (1) Peening is effective as a retrofit for shallow surface cracks that commonly occur at fillet weld toes. Peening imposes compressive residual stresses resulting from the plastic deformation induced by the peening hammer and reduces the geometrical stress concentration similar to that with grinding. Air hammer peening is effective in arresting fillet weld toe surface cracks with a depth of up to 3 mm (1/8 in.) if the tensile stress range does not exceed 40 MPa (6 ksi). Peening can also be applied to uncracked fillet welds to improve the fatigue resistance of the detail. The expected benefit of peening under favorable conditions (low stress range, low minimum stress) is an increase in fatigue life approximately equivalent to one fatigue design category (Fisher et al. 1979). (2) Peening should be done using a small pneumatic air hammer with all sharp edges of the peening tool ground smooth. Although peening intensity can be easily varied by changing air pressure, multiplepass peening at lower air pressures is most effective. Initial passes of the peening hammer may reveal some cracks that were not initially visible, and peening should be continued until the weld toe is smooth and no cracks are apparent. Penetrant inspection of the peened area should be conducted after peening to ensure that embedded flaws are not exposed. Peening is most effective when performed under dead load so that the imposed compressive residual stress has to be effective only against the live load. d. Gas tungsten arc remelting. (1) The GTA remelting process is also an effective procedure for repair of shallow surface cracks that occur at fillet weld toes. This procedure is generally effective for surface cracks with a depth of up to 8-7 EM 1110-2-6054 1 Dec 01 5 mm (3/16 in.) (slightly greater depths than peening) and is not limited to small stress ranges and minimum stress levels. Like peening, GTA remelting can also be used to improve the performance of uncracked fillet welds, approximately doubling the fatigue life. However, it is not as easily performed in the field, and it requires highly skilled welders and good accessibility. (2) With the GTA remelting procedure, a small volume of the weld toe and base metal is remelted with a gas-shielded tungsten electrode. After the area cools, the geometric stress concentration is improved and nonmetallic inclusions that might exist along the weld toe are eliminated. When the procedure is applied to crack repair, sufficient volume of the metal surrounding the crack must be melted so that upon solidification, the crack is eliminated. The effectiveness of the procedure is dependent on the depth of the remelted zone, since insufficient penetration would leave a crack buried below the surface. Such a crack would simply continue to propagate, resulting in premature failure. Proper selection of shielding gas and electrode cone angle is crucial in obtaining maximum penetration of the remelted zone. Argon-helium shielding and an electrode cone angle of 60 deg were found to be most effective (Fisher et al. 1979). For any retrofit procedure, the depth of penetration should be verified by metallographic examination of test plates before the procedure is applied in the field. 8-5. Rivet Replacement a. Missing, loose, or headless rivets and rivets with rosette heads should be replaced (Fazio and Fazio 1984). Deteriorated rivets missing more than 50 percent of the head should be replaced if the rivet is subject to an applied tensile force or tension resulting from prying action (Fazio and Fazio 1984). Rivet heads with rosettes and deteriorating projections should not be built up using weld metal or other materials (brazing, caulking), since these could aggravate rather than improve the condition. b. Rivets that require replacement should be replaced with high-strength bolts. However, removing a deteriorated rivet is sometimes difficult. The most accepted method of rivet removal is to knock off the rivet head using a pneumatic rivet buster and then force the rivet shaft out of its hole using a powered impact tool (Birk 1989). If necessary, the rivet hole should then be drilled out to obtain an aligned hole through the connected parts. Then a high-strength bolt is installed and tightened by an accepted method such as the turn-of-the-nut method. When pneumatic rivet busters are not available, rivet heads can be burned off. This technique can cause thermal metallurgical damage to the adjacent steel, and may result in burn gouges that adversely affect fatigue strength and susceptibility to corrosion. 8-6. Repair Examples a. Crack repair procedures developed for a cracked miter gate. (1) Description of condition. Figure 8-2 shows a crack in a tension flange of a girder on a miter gate (the photograph shows the inside face of the flange). The crack extends from the termination of a weld joining the flange of a diagonal bracing member and flange of the girder. Similar cracking occurred at perpendicular intersecting members (diaphragm and girder). Numerous through-thickness cracks similar to this occurred on the structure. (2) Cause of cracking. In general, cracking is attributed to high stress fatigue damage of low fatigue strength details. Cyclic loading occurs due to opening and closing of the gate leaves and to variation in hydrostatic pool. Unusually high stress may have occurred due to unintended loading while the gate was opened and closed with silt buildup at the gate bottom. Most of the cracks occurred at terminations of welds that join intersecting members, similar to the condition shown in Figure 8-2. Considering girder flexure, the fatigue strength of such details is Category E. 8-8