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.
8 improvement, capacity strengthening, and corrosion and impact damage repairs or retrofits. Fatigue Improvement This typically includes retrofit of out-of-plane and distortion- induced cracking by welding the connection stiffener to the flange. In most cases, the welding is on the tension flange. Welding of the transverse stiffeners and cross-brace connection plates to the tension flange was avoided prior to the early to mid-1980s because of concerns over fatigue cracking or brittle fracture of the tension flange (Zhao and Roddis 2004; Connor and Fisher 2007). As a result, a rigid connection between the girder flange and web-mounted stiffeners and connection plates was generally not provided. The lack of a positive connection often resulted in cracking from out-of- plane distortions, as shown in Figure 1, in a plane parallel to the primary loading stress (Zhao and Roddis 2004). In hindsight, not welding the connection plate to the tension flange led to out-of-plane cracking of the web within the small web gap between the web-to-flange weld and termination of the connection plate to web fillet weld, as shown in Figure 2 (Fish et al. 2015). At present, the current AASHTO LRFD design provisions (AASHTO 2014) require that the connection plate be welded or bolted to both the compression and tension flanges to resist out-of-plane cracking. Stiffening of the web gap by welding the connection plate to the tension flange is one of many common retrofit endorsements (Zhao and Roddis 2004). This retrofit strategy is simpler than a bolted retrofit; how- ever, it requires that the connection plate extend completely to the tension flange. Because the connection plates were often trimmed short of the tension flange, another plate may be required to be attached to the stiffener, which is then welded to the flange. This retrofit strategy also requires that the stress ranges in the flange not exceed the limits given by AASHTO Fatigue Detail Category C, which is the applicable detail category. If present, existing cracked welds are typically gouged out before the new welds are applied. It is important that finish grinding be performed after completion of welding to provide a smooth final surface. The use of field welding to repair out-of-plane and distortion-induced cracking is often not recommended by many DOTs as a result of the expen- sive labor necessary to ensure good weld quality and smooth surface grinding, along with the fear of introducing another INTRODUCTION The literature review process began with a search of the Trans- port Research Information Database (TRID) for documents related to field welding of highway bridges. The keywords and phrases that were searched included âfield welding,â âfield weld,â and ârepair weld.â Additional Internet searching was performed to find documentation from individual trans- portation agencies on field welding practices. The bridge and structural welding codes were also reviewed for any additional information or references. It was found that there is limited published research or literature on the topic of field welding and field repair welds. OVERVIEW Field welding is welding of a material outside of a fabrication shop. Field welding typically occurs at the bridge site. The emphasis of this synthesis is on planned field welded repairs and retrofits where consideration was given to design, speci- fications, procedures, qualifications, and inspection require- ments considered before performing work. EXTENT OF FIELD WELDING There is limited literature on the extent of field welding per- formed on steel bridges. Only one research document was found that specifically talked about the amount of field weld- ing performed in the bridge industry; NCHRP Report 321 (Gregory et al. 1989) found that most states would use welding as a repair method if it was proven in advance to be successful and if guidance was available on the subjects of inspection and quality control. They also reported that only one bridge owner claimed that they would never use welding for repair of cracks. According to this report, the agencyâs confidence appeared to be the key to repair welding because âThose states that did not like to weld in the field all stated that it was impossible to obtain welding and inspection personnel of a sufficiently high standard to produce a sound welded repairâ (Gregory et al. 1989). Types of Field Welded Repairs The literature review revealed that field welded repairs are performed on structures for three primary reasons: fatigue chapter two LITERATURE REVIEW
9 fatigue-sensitive detail in a primary load carrying member. Therefore, this method is often the last choice when other repair methods are not effective at stopping crack growth (Zhao and Roddis 2004; Hu et al. 2006). Hu et al. (2006) noted that the use of field welding to repair out-of-plane cracking, while convenient, should be minimized as a result of several potential problems. This includes overhead welding position, extensive cleaning to remove corrosion and dirt buildup, and preheat requirements that may be difficult with the large concrete mass above the top flange. In the 1970s, Fisher et al. studied the use of gas tungsten arc welding (GTAW), also known as tungsten inert gas (TIG) welding, to remelt the metal at the weld toe to improve the fatigue resistance. This method was only validated using cover plate termination details. In this process, a small volume of fillet weld and base metal is remelted by manually moving the tungsten electrode along the weld toe. This process can be used to improve the fatigue resistance at uncracked weld toes and to melt the volume containing existing shallow surface cracks to fully remove the crack. This research is documented in NCHRP Report 206 (Fisher et al. 1979). Capacity Strengthening This process typically involves a retrofit to increase capacity resulting from a poor load rating such as adding stiffeners to increase shear resistance. These repairs are undertaken in cases where the members were designed to support a lighter load than required under current specifications and are dis- tinguished from those where damage of some kind exists. Corrosion and Impact Damage Strengthening This includes repair or retrofit of damaged members. These members may have corrosion damage that has resulted in significant section loss and either requires additional stiff- ening or complete replacement of the member. Corrosion is typically repaired with the addition of reinforcing material or section replacement (Gregory et al. 1989). Impact damage frequently results in distorted and damaged members, which may also include cracking or tearing of the bridge member. Repair of impact damage normally involves straightening, with some welding or bolting of new or additional material (Gregory et al. 1989; Connor et al. 2008). NCHRP Report 321 (Gregory et al. 1989) found that the rate of the occurrence of these damage types varied from state to state. In some states, corrosion was the major problem with steel bridges, whereas others had major problems with accident damage and fatigue cracking. Field Welding Design Considerations Field welding presents special challenges when compared with original shop fabrication. Concerns regarding overhead positioning of many field welding repairs were raised in multiple sources (Miller 1993; Zhao and Roddis 2004; Hu et al. 2006). Another challenge is the weldability of the steel, because it may be of unknown composition. According to ANSI/AWS A3.0-2010, weldability is defined as the capacity of material to be welded under the imposed fabrication condi- tions into a specific, suitably designed structure performing satisfactorily in the intended service. Therefore, the weldability of a material for field welding must account for any adverse fabrication conditions that are often present in the field. FIGURE 1 Out-of-plane and distortion-induced cracking in web gap (adapted from Fisher and Keating 1989). FIGURE 2 Out-of-plane and distortion-induced cracking (Snyder 2015).
10 Field welded repairs and retrofits will likely involve the welding of types of steels where the ASTM specifications are out of date compared with current specified bridge steels; therefore, the current welding codes may not account for special requirements of these steels. When the composition of the steel is unknown or in doubt, it is necessary that analysis be performed so that a welding engineer can determine whether it is suitable for welding. It is imperative to know the chemical composition of the steel being joined (Miller 1993). Weld-access holes are used to provide access for repair welds in the flange of a girder and are placed in the web of the girder at the weld location. It is important that weld-access holes have adequate space to allow the welder to produce a good quality weld in this region. Run-off tabs and backing bars are utilized in welds requiring a CJP weld because this will help to eliminate the lack of penetration or lack of fusion from the welds. Generally run-off tabs and backing bars are taken off by grinding to remove any stress concentrations and to aid in inspection. Adverse fabrication conditions in the field include a sus- ceptibility to cracking when the steel base material is under high restraint. The rigidity of the structure or member does not allow for movement derived from shrinkage of the weld, which results in large shrinkage or residual stresses in the weld metal. Under conditions of high restraint, the likelihood of hydrogen cracks also increases, generally in the weld metal. This type of cracking may occur up to several hours after welding has been completed because the hydrogen in the weld metal will diffuse into the heat affected zone (HAZ) when the weld cools. Adequate preheat and welding sequences are the critical components to resisting weld cracking resulting from shrinkage stresses by slowing down the cooling rate and limiting the restraint. Other factors that affect hydrogen cracking are the thickness of the material and the type of joint because this influences the degree of restraint, the com- position of the steel, and the hydrogen content of the weld metal. The use of low hydrogen electrodes, which are dried at high temperatures before welding, can also aid in the preven- tion of hydrogen cracking. NCHRP Report 321 (Gregory et al. 1989) identified effec- tive field welding practices that can be used to improve the welded repair of cracks in steel members. It can be noted that this research was initiated before the publication of the first edition of the AASHTO/AWS D1.5 Bridge Welding Code. Therefore, any requirements in AASHTO/AWS D1.5 will take precedent over recommendations in this report. NCHRP Report 321 recommended the following: ⢠Because of the high levels of restraint typically present during field welded repairs and retrofits, minimum pre- heat temperatures of no less than 250°F should be used except when the steels are of quenched and tempered types. â For quenched and tempered steels such as ASTM A514 and A517, a separate table for minimum and maxi- mum preheat temperatures is provided. Overheating owing to high preheat and interpass temperatures and high heat input can possibly reduce the specified tensile and yield strengths, along with toughness. â The recommended minimum preheat temperature for ASTM A7 steel when carbon content is below 0.4% was 350°F. Because carbon content was not specified for ASTM A7 steels, a lower preheat and interpass temperature may be used for lower carbon contents. ⢠Whenever possible, preheating should be maintained throughout the duration of the repair. ⢠A preheating temperature of 100°F below the recom- mended preheat and interpass temperatures could be used before air-carbon arc gouging. ⢠To reduce the chance for delayed hydrogen cracking, steels with carbon content of greater than 0.4% may require additional heating immediately after welding is completed to maintain an elevated temperature for an extended period of time. ⢠Whenever possible, chemical or spectrographic analysis of a drilling or piece of steel should be performed even if the specification of the steel is known or mill test reports are available, because the steel may have been supplied out of specification or incorrect mill reports may have been supplied. This report included two tables for the ten- sile and chemical properties of the most commonly used steels in bridge construction over the previous 50 years before the publication of the report. This list does not include current ASTM specifications for bridge steels such as ASTM A709. CURRENT MANUALS AND SPECIFICATIONS FOR FIELD WELDING Two primary welding codes are used for structural welding of steel structures in the United States: AWS Bridge Weld- ing Code D1.5 (AASHTO/AWS 2010) and AWS Structural Welding Code D1.1 (AWS 2010b). AWS Bridge Welding Code D1.5 is required for new fabrication of bridges by the AASHTO bridge design specification. AWS Structural Weld- ing Code D1.1 is required for fabrication of buildings and other structures. The specific welding code to be applied to bridge field welding may vary from state to state. Each welding code contains requirements for the design of welds, qualification of welders and welding procedures, qualification of inspectors, and inspection and workmanship requirements. AWS Bridge Welding Code D1.5:2010 does not make a distinction between shop and field welds and does not con- tain any additional requirements for the strengthening and repairing of existing structures. When this code is specified on a field welding project, it is implied by the engineer that all of the code requirements for shop welds are also required for field welding including quality control requirements, having
11 qualified welding operators, written and qualified welding procedures, and workmanship requirements. AWS Structural Welding Code D1.1:2010, Section 8, on the other hand, includes additional requirements and guid- ance for field welding on existing structures. This section, âStrengthening and Repairing Existing Structures,â is quite brief, with only one and one-half pages of dialog. All of the requirements for shop welding elsewhere in the code apply to field welding on existing structures, along with the additional requirements in Section 8, which cover base metal, design of repairs and retrofits, fatigue life enhancements, workmanship and technique, and quality. AWS has published a guidance document for the strength- ening and repair of existing structures as AWS D1.7, Guide for Strengthening and Repairing Existing Structures (AWS 2010). Although this document is not a prescriptive code and does not include acceptance criteria, it does contain a considerable amount of useful information on what to consider before per- forming a field welded repair. This document contains exten- sive discussion on weldability of steel, including descriptions of the common classifications of structural steel, various car- bon equivalency equations, and common alloying elements. It also includes discussion on the means for determining an alter- nate acceptance criteria for discontinuities found on existing structures that do not meet the current fabrication codes. The qualified welding procedure has to meet the require- ments of the specified welding code and is based on mechan- ical and nondestructive test results of weldments. According to AWS D1.1 Structural Welding Code and AASHTO/AWS D1.5 Bridge Welding Code, welding procedures must be properly documented in a written format. This format, called a Welding Procedure Specification (WPS), may either be prequalified if all requirements for prequalification in the applicable welding code are met or may be qualified by testing in accordance with the requirements of the applicable code. Even if a WPS is prequalified, it must still be written and available to those authorized to use or examine them. In D1.5 Bridge Welding Code all procedures, except shielded metal arc welding (SMAW) that utilizes filler metals with a mini- mum specified yield strength less than or equal to 90 ksi, must be qualified by testing. In D1.1 Structural Welding Code, SMAW; submerged arc welding; gas metal arch welding, except short circuit transfer mode; and flux cored arc weld- ing processes may be prequalified if they meet the specified requirements. If the actual parameters of the field welding differ from the prequalification requirements, the procedures must be qualified through testing. A written Procedure Qualification Record is required under the AWS D1.1 and D1.5 welding codes to qualify a WPS through testing. The main parameters that are typically verified for accuracy are minimum preheat tempera- ture, minimum interpass temperature, welding current, voltage, travel speed, proper protection of filler metals, and appropri- ate welding techniques. The current, voltage, travel speed, and technique used to deposit the weld metal will control the heat input from welding that can affect the toughness of the base material, weld metal, and HAZ if it is not kept within acceptable limits. The ability of a welder to produce a weld of acceptable quality is verified through the use of welder performance qualification tests. Typically, the qualification lasts indefinitely provided that the welder carries out similar work in a certain time period without interruptions. However, this requirement may vary depending on specific agency requirements. Various state DOT manuals that cover field welding were reviewed to determine any aspects that are commonly included in state manuals. Welder qualification requirements were included in the Iowa (Iowa DOT Office of Materials 2011), North Carolina (North Carolina DOT Materials and Tests Unit 2006), New York (New York Office of Structures 2008), Ohio (Ohio Department of Transportation 2008), Oklahoma (Oklahoma DOT Materials & Research Division 2012), and Texas (Texas Department of Transportation 2004) DOT state manuals. Generally, welder qualification programs are insti- tuted by each state DOT where the welder qualification tests are conducted or witnessed by state representatives and a record of field welders who have successfully passed the qualification test are entered into the departmentâs database. These DOTs require that all personnel performing field welding on their construction project pass the welder qualification process and be documented as qualified welders. In AWS D1.5, the welder must be qualified for the position in which they will weld, and a table was developed that lists the qualified positions for each test position. Oklahoma DOT requires that all bridge welders test on a fillet weld assembly in the vertical (3F) and overhead (4F) positions, along with groove weld assemblies in the horizontal (2G), vertical (3G), and overhead (4G) positions (Oklahoma DOT Materials & Research Division 2012). North Carolina DOT requirements are very similar to Oklahoma DOT except they do not require the horizontal (2G) position (North Carolina DOT Materials and Tests Unit 2006). Texas DOT requires that bridge welders who perform groove welding pass the qualification test for groove welds for plates in the vertical (3G) and overhead (4G) positions along with other additional requirements (Texas Department of Transportation 2004). New York DOT requires that bridge welders who perform groove welding pass the qualification test using either the vertical (3G) and/or overhead (4G) positions depending on the project required positions (New York Office of Struc- tures 2008). The Oklahoma, North Carolina, and Texas DOT requirements are tougher than the requirements in AASHTO/ AWS D1.5. Iowa DOT includes the same table for the quali- fied positions as AASHTO/AWS D1.5 (Iowa DOT Office of Materials 2011), and the Ohio DOT manual states that the welders must be qualified for the position in which they are welding (Ohio Department of Transportation 2008).
12 It was also found that many of the state agencies had field welding inspection guides that included information to aid weld inspectors. This includes, but is not limited to, information on the typical weld symbols, welding procedures, electrodes and electrode storage, preheat and interpass temperatures, weld joint preparation and cleaning, weld inspections, equipment, and typical weld discontinuities. An example of good and bad weld beads from Oklahoma DOT is shown in Figure 3. QUALITY CONTROL AND QUALITY ASSURANCE One of the greatest concerns is the quality of the field welds. Field welding requires the same quality control as shop weld- ing; however, quality control is more difficult to maintain under field conditions. During the literature review, this concern was expressed in multiple references (Gregory et al. 1989; Miller 1993; Keating et al. 1996; Zhao and Roddis 2004). Oklahoma DOT Examples of Good and Bad Beads Shielded Metal Arc Welding (SMAW) FIGURE 3 Oklahoma DOT weld quality examples (Oklahoma DOT Materials & Research Division 2004).
13 This does not imply that the quality or quality control requirements are different for field welds than for shop weld- ing. It simply means that it can be more difficult to maintain quality control under field conditions. The control of field welded repairs often relies heavily on the competency of the contractor performing this work. Engineers tend to rely on the contractorâs expertise, experience, and reputation to ensure quality workmanship. Even with an experienced contractor, the engineer and quality assurance inspectors must ensure that they responsibly perform their duties. Welder Qualification According to the AWS D1.1 and 1.5 welding codes, welding should only be permitted by individuals who have passed the appropriate qualification tests in accordance with the provi- sions of the specified code. According to NCHRP Report 321 (Gregory et al. 1989), the WPS and any applicable Proce- dure Qualification Record should be subject to the approval of the Engineer of Record for the repair welding before the welding is conducted. This report goes on to state that a sim- ple test, such as the fillet weld break test, may be required when the quality of the welderâs work does not appear to be up to the required standard. This report also advised that the welder carry out a practice repair weld at the actual site of the repair that duplicates the conditions that will be encountered. It is important that this rehearsal of the repair include appro- priate plate thickness and arrangement of the plates to repro- duce the access available on the actual bridge. The practice repair provides an opportunity to verify that the welder is suf- ficiently prepared and that the equipment is adjusted properly before beginning the bridge welding. The rehearsal repair could include mock-up cracks marked on the plates and any preheat, removal, and post-heat requirements. These welds can be evaluated using NDT to verify their quality before the actual repair welding is initiated. In addition, mechanical properties may be required for welding on a fracture critical member (FCM). Weld Inspection It is important that the inspection of repair welds be at least as thorough as when the bridge was originally fabricated (Gregory et al. 1989). This includes matching the current requirements for the methods and extent of the inspection com- pared with current bridge fabrication practices. Acceptance requirements for visual and NDT of the welds are included in the AWS welding codes and, typically, the require ments for AWS D1.1 or AASHTO/AWS D1.5 will be specified for each project. The acceptance criteria in the AASHTO/AWS welding codes include the size and distribution of welding discontinuities that are either permissible or rejectable. The requirements in AWS D1.1 may be modified by the engineer through the project specifications and any additional require- ments can be specified. Although unlikely, the engineer may also decide to use a fitness-for-service approach to specify the acceptance criteria of field welds. Typically, the requirements of BS 7910 (British Standards 2013) or API 579 (American Petroleum Institute 2007) are applied to a fitness-for-service evaluation of steel structures. AWS D1.1 and D1.5 welding codes require that welding inspectors be suitably qualified to one of the following stan- dards and that this qualification be documented by: ⢠Current or previous certification as an AWS Certified Weld Inspector (CWI) in conformance with the provi- sions of AWS QC1, Standard for AWS Certification of Welding Inspectors. ⢠Current or previous qualification by the Canadian Weld- ing Bureau in conformance with the requirements of the Canadian Standard Association Standard W178.2, Certification of Welding Inspectors. ⢠An individual who, by training or experience, or both, in metals fabrication, inspection, and testing, is competent to perform inspection of the work. Visual testing welding inspection is the most important examination method and is applied to all repair welds, inde- pendent of the utilization of any other inspection methods (Gregory et al. 1989). Welding inspectors are provided with the repair procedure and details, drawings or sketches, and acceptance criteria. Visual inspection is typically performed before welding, during welding, and after welding. Before welding, visual inspection includes confirmation of groove dimensions, cleanliness and surface finish of the groove, removal of any existing weld discontinuities, fit-up of plates, and fit-up of backing bars (if present). During weld- ing, visual inspection includes confirmation of preheat and interpass temperatures, welding sequence, electrode require- ments, storage and handling, welding variables, and post-weld heat treatment. After welding, visual inspection includes con- firmation of backing bar removal, surface finish, size of welds, and lack of rejectable weld discontinuities. Typical NDT methods applied to field welded repairs and retrofits include: ⢠Dye penetrant testing (PT), ⢠Magnetic particle testing (MT), ⢠Ultrasonic testing (UT), and ⢠Radiographic testing (RT). PT is a convenient and simple NDT method that is lim- ited to inspection of surface discontinuities. High or low temperatures can affect the results of PT; therefore, NCHRP Report 321 advised against using this method when the steel temperature range is not within 40°Fâ110°F (Gregory et al. 1989). MT is limited to inspection of surface or near surface discontinuities in ferromagnetic materials. This method can
14 be slightly more sensitive to tightly closed cracks; however, it can be more time consuming and requires more skill to differentiate actual discontinuities from false calls. MT can be affected by wind. Both UT and RT can be used to inspect for internal dis- continuities. These methods require advanced training to properly perform the inspection and evaluate the results. Radiography in the field can have issues resulting from safety concerns and portability. Large thicknesses may require different sources that may be less portable than iridium sources that can be used on thinner material. UT can also be affected by temperature because the couplant can either dry out if the steel is very hot or freeze if the steel is too cold. The AWS D1.1 and D1.5 welding codes require that per- sonnel performing NDT other than visual be qualified in accordance with the current edition of the American Society for Nondestructive Testing, Recommended Practice Number SNT-TC-1A as NDT Level II or as NDT Level I working under the NDT Level II. Welding Cleanliness Adequate cleaning of the field weld location to remove paint, galvanizing, dirt, loose or thick scale, slag, rust, moisture, grease, or other impurities is required by the AWS welding codes. Scaling rust, pits, or other surface irregularities can affect both the welding and subsequent NDT. Oklahoma DOT (Oklahoma DOT Materials & Research Division 2004) states that one of the most practical tools to clean a weld joint is a stiff wire brush. AWS welding codes allow mill scale that can withstand vigorous wire brushing to remain. According to the welding code, all finished welds must be cleaned as well and recommends that multiple pass welds be cleaned between every pass (Oklahoma DOT Materials & Research Division 2004). Typically, this is accomplished using a chipping hammer and stiff wire brush. A grinder may also be used with care for cleaning, but be used to avoid doing more harm than good on the finished weld and base metal. Risks associated with grinding include excessive removal of base metal and blemishing over weld defects. Slag, which is a byproduct of some welding processes and protects the weld when it is molten, must be removed between passes and at completion of the weld. Failure to clean the surfaces before welding may encourage the formation of porosity in the welds or lack of fusion between the weld and parent material (Miller 1993). If a project will involve repairing a discontinuity in an existing weld, it is essential that the discontinuity be com- pletely and fully removed prior to the subsequent weld- ing. Grinding the location to bright metal and performing NDT using PT or MT is ideal to ensure complete removal (Miller 1993). The cleanliness of the materials to be welded is frequently a problem when performing a field repair (Miller 1993). According to the welding codes, loose scale and rust must be removed, at a minimum, before welding. Wire brushing, shot or sandblasting, or grinding are the typical methods used to clean the steel. AWS D1.1 and D1.5 welding codes require that welding consumables that have been removed from the original pack- age be protected and stored so that the welding properties are not affected. NCHRP Report 321 reported that it is important that the welding electrodes are dried at high temperatures before use on bridge repair to ensure that the weld metal will have extra low hydrogen content (Gregory et al. 1989). The extra low hydrogen content (of less than 5 ml/100 g) is obtained when low hydrogen electrodes are taken from hermetically sealed containers and dried at 700°F to 800°F for 1 h and used within 2 h after removal. Heated ovens are required by the AWS welding codes to store low hydrogen electrodes after the hermetically sealed containers are opened in order to maintain their low hydro- gen characteristics. Because of the limited electrical power in the field, especially for overnight storage, purchasing elec- trodes in a small container, such as a 10-lb, hermetically sealed container, may preclude the need for drying ovens if the unused electrodes are thrown away after opening. Coiled electrodes must be protected from direct contact with moisture and condensation. Weld Fitup Fitup may be difficult for field welding because of the inability to fixture the pieces or limited access. Field welds are to be designed with consideration of the access and visibility for the welder to perform a quality weld. Welding out of the typi- cal position is likely because the structure cannot be moved to a position of greatest advantage. It is best to select weld joints that will minimize the amount of out-of-position welding (Miller 1993). Welding Environment Although the ambient environment where field welding occurs is typically not as controlled as the environment in a bridge fabrication shop, protection from wind and moisture can be provided by constructing a temporary enclosure around the welding location. Control of the ambient temperature in the immediate area surrounding the point of welding may be required if the tem- perature is below the limits in the welding code. AWS D1.1 and D1.5 require that the steel be preheated to 70°F if the ambient temperature drops below 32°F. Any additional requirements for preheat beyond this amount will still be required. Welding is not to be performed at an ambient temperature below 0°F
15 in the immediate vicinity around the weld. It is suggested by Miller (1993) that the performance of the welders is negatively affected at such low temperatures and, hence, the practice is discouraged. Protection from low temperatures can be pro- vided by erecting a tent or other protection around the welding location. In this instance, the temperature requirements refer to the temperature within the enclosure at the welding location and not necessarily the temperature outside of the enclosure. It can also be noted that problems with equipment may occur at such low temperatures. For example, the lubricants in the wire feeder gear boxes may thicken at low ambient tempera- tures, which could adversely affect delivery of the electrode to the arc. NCHRP Report 321 suggests that electric resistance heat- ing elements with thermocouples and automatic temperature controllers and recorders be used when close control of pre- heating temperatures is required or when extensive repairs are being performed (Gregory et al. 1989). Moisture must be controlled during field welding. The base metal must be dry before weld metal is deposited. Pre- heating the plates before welding can be used to ensure that they are dry; however, an enclosure or cover may be necessary to ensure that they stay dry. Wind can cause problems with gas arc shielding because it will not allow for adequate coverage of the weld during the welding process. Gas metal arc and gas-shielded flux cored arc welding processes are most sensitive to this issue (Miller 1993). The use of self-shielding flux core or SMAW is often preferred for field welding to avoid problems caused by wind. Where gas-shielded processes are required, a wind screen would be used to protect the gas shielding. PERFORMANCE OF REPAIRS AND RETROFITS Little research or published literature was found that specifi- cally documented the performance of field welding, likely because the only inherent difference between field welding and shop welding is where the welding takes place. The weld- ing processes, welding parameters, inspection, and quality requirements are generally the same for field welding as they are for shop welding. One difference between fabrication shop and field welding is the additional restraint that may be present in a field welded repair or retrofit owing to the structural connections between bridge members. Another difference is that field welding repair and retrofits may be performed while under active live loading, which can cause vibration along with additional dead loading. This concern was mentioned by Zhao and Roddis (2004) and Hu et al. (2006) because micro-cracking could occur within the HAZ as a result of structure vibration dur- ing solidification from repairs carried out under traffic. Two research studies were found that attempted to capture these effects by performing experimental testing of fatigue crack weld repair while under tensile stress or dynamic loading. NCHRP Report 321 described experimental work on repair welded fatigue cracks (Gregory et al. 1989). This research included fatigue testing girders until cracks grew to a pre- determined length and then repair welding the cracks before performing additional fatigue testing. The tests demonstrated that repair welds had at least the fatigue life as the original shop weld. Fracture toughness testing was performed that demonstrated that multiple repair welds could be made with minimal toughness reductions in the HAZ. The repair welds met the relevant AASHTO requirements for fracture tough- ness; however, it was suggested that repairs on FCMs include mechanical testing to qualify the welding procedures and ver- ify that the increased toughness requirements are met. Experimental work was undertaken to determine the effect of repair welding under dynamic loading when the crack may be opening and closing. It was found that it is possible to carry out repair welding under dynamic loading, but that the preferred procedure was to close the bridge to traffic while the root pass and possibly a second pass of weld metal are deposited. It was determined that welding is an effective and economic method for the repair of fatigue cracks if a good quality weld can be guaranteed. It was stressed that the pres- ence of weld defects can severely reduce the fatigue strength of repair welds. Another study of the performance of field welded repairs on steel highway bridges was performed in 1984 by Matsumoto and Motomura (1984). This research evaluated the effect of field welding under the influence of load-induced stresses and traffic-induced vibrations that may be present when field welding is done while the bridge is open to traffic. Experimen- tal tests were performed on plate girder bridges to simulate these effects while rewelding fatigue cracks. The tests were done on ASTM A572 Grade 50 and ASTM A678 Grade B steel. This research found that the influence of low tensile stresses on fillet weld defects was almost negligible and that fillet welding can be applied to repair work performed under the influence of tensile stresses. Fillet welding was per- formed on web plates and flange plates under the influence of vibration. It was concluded that fillet welding can be used to repair web plates and flange plates under the influence of vibration if visual examination and the correction of defects are performed carefully. Repair procedures for out-of-plane and distortion-induced fatigue cracking utilizing gouging and rewelding of cracks were evaluated by Keating et al. in 1996. This study included field investigation of repairs performed on an in-service bridge, along with laboratory fatigue testing of repairs on large-scale specimens and finite-element analyses. This method was found to be a viable option for the repair of out-of-plane and distortion-induced fatigue cracking as long as proper proce- dures were followed to achieve a quality repair. The gouging
16 and rewelding technique required a high degree of skill and inspection and, therefore, expense. This type of repair restored the cracked location to its uncracked state; however, without an additional retrofit to lower or eliminate the out-of-plane stresses in this region, the crack would likely reappear. An in-depth inspection including NDT can help ensure complete removal of the crack. As part of this study, field welding was also performed to connect the existing connection plates to the flange to pro- vide a rigid connection. Concerns with this retrofit included the addition of a fatigue-sensitive detail and the quality of the field welding. Concerns also arose over the inability to adequately clean the joint before welding if the connection plate is tight fit to the flange. Additional grinding may be required to ensure that no paint is in the tight-fit gap. NCHRP Report 604: Heat-Straightening Repair of Dam- aged Steel Bridge Girders: Fatigue and Fracture Performance (Connor et al. 2008) investigated the effect of multiple dam- age repair cycles on the fatigue and fracture performance of steel girders. During this study, many girders required repair welds as a result of cracking induced from impact damage applied in the lab. The impacts were all applied dynamically using a large drop-weight type system. Many of the repairs were significant because of tearing at stiffeners and cover plate details. It was demonstrated that after heat straightening and placement of the repair welds, the repaired welded details did not perform any differently during fatigue testing when compared with the original as-fabricated welds. Although the weld repairs were made in the laboratory, the authors indi- cated that as long as the repair welds are performed properly and utilize sound welding procedures, there should be no concern as to the performance of the repair welds in terms of fatigue or fracture limit states. Kelly and Dexter (1997) researched the fatigue perfor- mance of repair welds on details commonly implemented on ship structures. Fatigue tests were conducted on full-scale welded beams with a variety of butt welds in the flanges and weld-access holes in the webs. These welds would not meet AASHTO Category B requirements because some of the welds did not have the weld reinforcement ground smooth, backing bars removed, or were welded on one side without backing. The results showed that the fatigue strength of these butt welds corresponded to the AASHTO Category D S-N curve. This fatigue strength was not affected by the type of steel, whether they were two-sided or one-sided welds, whether backing bars were present or removed, and with and without the edges ground flush. Weld-access holes required to make repairs may also be characterized as Category D details. The research included multiple repairs at the same detail by fatigue testing until cracking occurred between each repair cycle. The fatigue testing indicated that the weld repairs of through-thickness cracks have the same fatigue strength as the original new butt welds, even after repairing the same location up to four times. Fisher et al. (1979) performed GTAW to remelt the weld toes of precracked cover plate termination as reported in NCHRP Report 206. It was found that the proper selection of shield- ing gas and electrode cone angle was necessary to provide adequate penetration to completely remelt the cracks. Mill scale was removed before welding because it was found that undercutting would occur if the scale was not removed. Over- head welding was undertaken to simulate the field conditions for cover plate terminations. It was found that a several hours of training were required before welding personnel could achieve the desired retrofit condition. This research found that GTAW remelting was a successful method for repairing weld toes that have a crack growth less than 0.180 in. deep and 3 in. long, provided that adequate penetration is provided and the operator performs a proper weld. It was also deter- mined that this method could be provided in field conditions under normal traffic loading. GTAW (also known as TIG) remelting of weld toes to repair fatigue cracks of fillet welds was also investigated by researchers in Japan (Natori et al. 1989). They had findings similar to that of Fisher that the vibration of the bridge under normal service conditions did not affect the GTAW remelt. They recommended the use of GTAW remelting as an effective retrofit technique for cracks less than 2 mm (0.08 in.) deep. Using an assumed crack depth to length ratio of 1â5, it was estimated that cracks of up to 10 mm (0.39 in.) in length can be removed by the GTAW remelting process. Fisher et al. (1982) recommended the use of field welding to retrofit an out-of-plane cracking detail by welding a shear tab to the transverse connection plate and flange of curved, continuous box girder bridges in Baltimore, Maryland. This retrofit was determined to be the only conceivable possibility in the negative moment region, because cutting back the con- nection plates to increase the web gap length and provide additional flexibility would likely alter the structural behav- ior and result in increased adverse behavior. The retrofit was done in 1982, but cracking was observed in the welds dur- ing a site inspection in 1984. It was later reported that the cracking was the result of undersized and poor quality welds (Demers and Fisher 1990). In 1986, the shear tabs were par- tially removed and retrofit holes were drilled in the crack tips. No additional cracks were reported. Koob et al. (1985) investigated the cracking that occurred on the Poplar Street Bridge in St. Louis, Missouri. It was deter- mined that this cracking occurred because of out-of-plane and distortion-induced fatigue. Various retrofit strategies were investigated, including field welding the connection plates to the top flange of the girder. The main concerns for this repair were the weld quality of overhead welding under the field conditions and the effect of traffic during welding. However, it was determined that a welded retrofit was worthy of evalua- tion owing to the ease of installation and the good performance record of welded connection plates. Because of concerns
17 about micro-cracking in the HAZ during solidification when welding under live loading, a decision was made to stop traf- fic during the welding operation for the trial retrofits. Slight undercutting was noted in the top flange weld toe and debris between the stiffener and the top flange caused impurities in the root pass. Poor quality locations were ground out and rewelded. It was determined that sand blasting may not be the most effective method for paint removal because this might have contributed to the trapped debris. A carbide burr grinder was used to weld the toe ground to provide a smooth transition; however, weld toe peening was not done in the trial retrofit. This retrofit was found to eliminate out-of-plane displacement of the web gap; a trial bolted retrofit was also tested in this study and not found to eliminate out-of-plane displacement of the web gap. In the end, a softening technique where the web gap is increased to decrease the stresses was recommended over the field welded retrofit because of con- cerns over the quality of the field welds. Kansas DOT performed field welding to repair out-of-plane cracking in the Westgate Bridge, a two-girder bridge consisting of a girder/truss floorbeam/stringer system (Zhao and Roddis 2004). The two girders are 1,758 ft long and are fabricated from A36 material. The bridge was built in 1977, and by 1994 nine of the 11 girder spans had developed horizontal or horseshoe cracks at the interior floorbeam to girder connec- tions. Along with 1-in. stop crack holes being drilled at the end of the cracks, field welding was performed to repair the cracking. In positive moment regions, the web gap locations were stiffened by placing 5â16-in. fillet welds to connect the stiffeners to the top flanges. In negative moment regions, 3â4 in. stiffener plates were added on the other side of the girders to resist the out-of-plane distortion. Field welds were placed to weld the existing stiffeners to the bottom (compression) flange; however, no welding was placed on the top (tension) flange. The new stiffeners were tight fit to the top flange and cut short where they intersected with existing longitudinal stiffeners. No new crack development has been reported since the bridge was repaired. Cracking on the Blanchette Bridge over the Missouri River, owned by the Missouri DOT, was noted on stringers after the concrete deck was replaced with a steel grid deck (Marianos et al. 2006). This cracking stemmed from field welds that connected the shim plates of the grid deck to the top flange of the stringers. The stringers conformed to the requirements of ASTM A7 steel, whereas the shim plates were ASTM A36 steel. ASTM A7 steel is generally considered weldable, and no previous indication of welding issues was found during inspection of the fillet welds. An in-depth study was performed by Missouri DOT to determine the cause of cracking. It was determined that the cracking was the result of open shim butt joints where the shim plate was not spliced before being fil- let welded to the stringers. This type of cracking would have occurred independent of whether the fillet weld was performed in the field or the shop. SUMMARY The literature review revealed that field welded repairs are performed on structures for three primary reasons: fatigue improvement, capacity strengthening, and corrosion and impact damage repairs or retrofits. ⢠Fatigue improvement typically includes retrofit of out- of-plane and distortion-induced cracking by welding the connection stiffener to the flange. In most cases, the welding is on the tension flange. Along with field welded retrofits for out-of-plane cracking, GTAW or TIG weld- ing, to remelts the metal at the weld toe to improve the fatigue resistance. ⢠Capacity strengthening typically involves a retrofit to increase capacity resulting from a poor load rating such as adding stiffeners to increase shear resistance. These repairs are used in cases where the members were designed to support a lighter load than required under current specifications and are distinguished from those where damage of some kind exists. ⢠Corrosion and impact damage strengthening include repair or retrofit of damaged members. Such members may have corrosion damage that has resulted in significant section loss and either requires additional stiffening or complete replacement. Two primary welding codes are used for structural welding of steel structures in the United States: ⢠AWS Bridge Welding Code D1.5 is required for new fabrication of bridges by the AASHTO bridge design specification. ⢠AWS Structural Welding Code D1.1 is required for fabrication of buildings and other structures. The specific welding code to be applied to bridge field welding may vary from state to state. AWS has published a guidance document for the strength- ening and repair of existing structures as AWS D1.7 Guide for Strengthening and Repairing Existing Structures. Although this document is not a prescriptive code and does not include acceptance criteria, it does contain useful information on things to consider before performing a field welded repair. Many state DOTs sponsor welder qualification programs where the welder qualification tests are conducted or wit- nessed by state representatives and a record of field welders who have successfully passed the qualification test can be found in the departmentâs database. These DOTs require that all personnel performing field welding on construction projects pass the welder qualification process and be documented as qualified welders. Many of the state agencies had field weld- ing inspection guides that included information to aid weld inspectors. The quality of the field welds is a concern that was raised in multiple references. Quality control is more difficult to
18 maintain under field conditions; however, the quality require- ments are the same for field welds as shop welds. Research on field welding has been performed using experimental testing of fatigue crack weld repair while under tensile stress or dynamic loading. These studies found that it is possible to carry out repair welding under dynamic load- ing; however, the preferred procedure is to close the bridge to traffic while the root pass and possibly a second pass of weld metal are deposited. Other research performed on the fatigue resistance of field welds found that repair welds had at least the fatigue life of the original shop weld, as long as proper procedures were followed to achieve a quality repair. The repair welds met the relevant AASHTO requirements for fracture toughness; however, it was suggested that repairs on FCMs include mechanical testing to qualify the welding procedures and verify that the increased toughness requirements are met. Research has also demonstrated that after heat straightening and placement of the repair welds the repaired welded details did not perform any differently during fatigue testing when compared with the original as-fabricated welds. Field welded repairs and retrofits that have been imple- mented on steel bridges and documented in literature have performed well in service. Only one issue with a field welded repair or retrofit was found during the literature review and this was attributed to undersized and poor quality welds.
