DEPARTMENT OF THE ARMY ETL 1110-2-346U.S. Army Corps of Engineers

CECW-ED Washington, DC 20314-1000

TechnicalLetter No. 1110-2-346 30 September 1993

Engineering and DesignSTRUCTURAL INSPECTION AND EVALUATION

OF EXISTING WELDED LOCK GATES

1. Purpose

This engineer technical letter (ETL) providesguidance for evaluating the structural adequacy ofexisting welded lock gates.

2. Applicability

This ETL applies to HQUSACE elements, majorsubordinate commands, districts, laboratories, andfield operating activities having responsibilities forcivil works projects.

3. References

a. ER 1110-2-100, Periodic Inspection andContinuing Evaluation of Completed Civil WorksStructures.

b. ER 1110-2-101, Reporting Evidence of Dis-tress of Civil Works Projects.

c. EM 1110-2-2105, Design of Hydraulic SteelStructures.

d. EM 1110-2-2703, Lock Gates and OperatingEquipment.

4. Discussion

a. ER 1110-2-100 defines periodic inspectionrequirements for completed civil works projects.These requirements include all aspects of a projectand are general in nature. ER 1110-2-101 requires

that signs of distress in any project feature bereported through channels to HQUSACE. Neitherof these references describes how to perform adetailed inspection and evaluation of hydraulic steelstructures.

b. The state of the art in metal fatigue andfracture analysis has advanced greatly in recentyears. In many industries these concepts are regu-larly applied to new designs and to evaluation ofexisting structural elements. EM 1110-2-2105requires that fatigue and fracture be consideredwhen designing new hydraulic steel structures.

c. Steel structures at several civil works pro-jects have experienced severe cracking. Some ofthese incidents are discussed in Enclosure 1. Thisdemonstrates the need to emphasize fatigue andfracture concepts when inspecting and evaluatingsuch structures.

d. The six enclosures to this ETL providedetailed methods for inspection and evaluation ofexisting steel lock gates. These enclosures providespecific recommendations for inspection techniques,evaluation of detected flaws, and prediction ofremaining life. These concepts are also applicableto a wide range of other structures, including almostany steel structure in a civil works project.

5. Action

a. Periodic inspection of steel lock gatesshould include close visual inspection of criticalmembers and connections.

ETL 1110-2-34630 Sep 93

b. If cracks are detected during periodicinspections, the cracked elements and other criticallocations should be evaluated using the methodsdefined in the enclosure.

c. These actions should also be implementedfor other steel features of civil works projects whendeemed appropriate by the structural engineer.

FOR THE DIRECTOR OF CIVIL WORKS:

6 Encl PAUL D. BARBER, P.E.Chief, Engineering DivisionDirectorate of Civil Works

2

ETL 1110-2-34630 Sep 93

GENERAL DISCUSSION

1. Scope

a. Enclosures 1 through 6 include procedures toinspect existing welded steel lock gates and evaluatethem for potential failure. The general conceptsmay also be applied to new designs, riveted andbolted gates, gates for other purposes, and even toother types of materials.

b. This enclosure provides general discussionand Enclosure 2 discusses causes of structural dete-rioration. Enclosure 3 describes the level of steelgate inspection appropriate during a periodic inspec-tion. This includes preselecting critical locationswhich require close examination, including identifi-cation of fracture critical members and connectionsand visual inspection. Enclosure 4 describes thedetailed nondestructive testing procedures whichshould be used while performing a detailed struc-tural inspection. Some of these procedures mayalso be appropriate during periodic inspections.

c. When evaluating older lock gates, necessarymaterial information may not be available. It maybecome necessary to perform material testing todetermine the chemistry, strength, ductility, hard-ness, and toughness of the base and weld metal.For this reason, material and weld testing tech-niques are discussed in detail in Enclosure 5.

d. Engineering evaluation of an existing gateshould be more than an educated guess or a subjec-tive evaluation. The gate condition should be deter-mined numerically using proper fatigue and fractureanalysis methods. These procedures are describedin Enclosure 6. The analyses can be used to deter-mine if the gate is safe to continue currentoperation, what is a safe interval until the nextinspection, and what is the remaining life of thegate for expected operating conditions.

2. Types of Gates

a. Currently, the U.S. Army Corps of Engineers(USACE) operates over 250 lock chambers. Thefunctional requirements for lock gates vary, depend-ing on the specific project location and operatingconditions. The primary purpose for steel gates isto provide a damming surface across the lock

chamber; however, they can also be used as guardgates, valves for filling and emptying the lockchamber, for passing ice and debris, to unwater thelock chamber, to separate salt and fresh water, andto provide access from one lock wall to the othervia walkways attached to the top of the gates. Mostexisting lock gates are miter gates and vertical-liftgates, with a small percentage being sector gatesand submergible tainter gates.

b. The majority of lock gates are of the mitertype, primarily because they tend to be more eco-nomical to construct and operate and can be openedand closed more rapidly than other types of lockgates. Miter gates are categorized by their framingmechanism into vertically or horizontally framedgates. Water pressure acting on the skin plate of avertically framed gate is resisted by vertical beammembers supported by a horizontal girder at the topand bottom of the leaf. The horizontal girders thentransmit the loads to the miter and quoin at the topof the leaf and into the sill at the bottom of the leaf.Horizontally framed lock gates transmit the skinplate water load directly to horizontal girders whichthen transfer the load to the quoin block and intothe walls of the lock monolith. Current designguidance, Engineer Manual (EM) 1110-2-2703,"Lock Gates and Operating Equipment,"1 recom-mends that future miter gates be horizontallyframed; however, a large percentage of existingmiter gates are vertically framed.

c. Another type of lock gate is the sector gate.This gate is framed similar to a tainter gate, how-ever, it pivots about a vertical axis similar to amiter gate. Sector gates have traditionally beenused in tidal reaches of rivers or canals and conse-quently may be subject to head reversal. Sectorgates may be used to control flow in the lock cham-ber during normal operation or close off flow dur-ing emergency operation. Sector gates are generallylimited to lifts of 10 ft or less.

d. Vertical-lift gates differ from miter andsector gates in that they are raised and lowered

____________________1 References for Enclosures 1 through 6 are foundin the Reference section following Enclosure 6.

Enclosure 1 1-1

ETL 1110-2-34630 Sep 93

vertically to open or close the lock chamber. Theload developed by water pressure acting on the liftgate skin plate is transmitted along horizontalgirders into the walls of the lock monolith. Liftgates can be operated under moderate heads but notunder reverse head conditions.

e. Submergible tainter gates are used infre-quently as lock gates. This type of gate pivotssimilar to a spillway tainter gate but is raised toclose the lock chamber and lowered into the cham-ber to open it. The load developed by water pres-sure acting on the submergible tainter gate skinplate is transmitted along horizontal girders to strutsrecessed in the lock wall. The struts are connectedto and rotate about trunnions anchored to each lockwall.

3. Strength and Serviceability Requirements

a. Lock gates are designed according torequirements of appropriate EM’s and design codesas listed in EM 1110-2-2105, "Design of HydraulicSteel Structures." Lock gates are designed to havedesign strengths at all sections equal, at least, to therequired strengths calculated for the critical combi-nation of loads and forces. Various gate membersmust be designed to resist axial forces, bendingforces, and combined bending and axial forces.These members are fabricated from bars, plates,standard rolled shapes, and built-up sectionsdepending on geometrical requirements, loading,and economics. Structural inspection and evalua-tion are required to assure that adequate strengthand serviceability are maintained at all sectionsduring the life of the gate.

b. Serviceability is a state in which the functionof a lock gate, its maintainability, durability, andoperability are preserved for the life of the gate.The structural inspection and evaluation must assurethat all deflections, deformations, vibrations, corro-sion, and wear of structural members do not impairthe operability or performance of the lock gate.

4. Examples of Distressed Lock Gates

Fracture and failure of steel members and connec-tions have occurred in several Corps of Engineersprojects. These projects received the required peri-odic inspections. However, the inspections were

not detailed enough to detect initial cracks nor werefatigue and fracture analyses performed for thesestructures prior to, and often not subsequent to,failure. The following brief examples, all takenfrom a single district, illustrate the potential resultsof casual inspection combined with inattention tofatigue and fracture concepts during design.

a. Miter gate anchorage.

(1) The project utilized vertically frameddownstream miter gates, 45 ft high, with a 110-ftlock width. The upper embedded gate anchoragefailed unexpectedly while the chamber was at tail-water elevation. Failure occurred by fracture at thegudgeon pin hole. The anchor was a structural steelassembly of two channels and two 1/2-in.-thickplates. The use of a channel with up-turned legscauses ponding of water and results in pitting andscaling corrosion. Since the anchor is a nonredun-dant tension member, failure caused the leaf to fallto the concrete sill, though it remained vertical.

(2) The failure surfaces were disposed ofwithout an examination to determine the cause offailure. To make the lock operational as quickly aspossible, repairs were implemented without anyevaluation or recommendations from the Engineer-ing Division. These repairs consisted of butting andwelding a new channel section to the remainingembedded section and bolting a 1-in. cover plate tothe channel webs. The bolt and plate materials arenot known.

(3) The same type of anchorage is used on atleast two other projects with a total of 16 similaranchors.

b. Spare miter gate.

(1) The project had a spare miter gate whichconsisted of five welded modules. When in use,these modules were stacked vertically and boltedtogether. The spare gate had been used severaltimes. However, 1 month after the last installation,cracks were discovered in the downstream flangesof three vertical girders. The cracks originated atthe downstream face of the flange in the heataffected zone at the toe of a transverse fillet weld.(This detail is category E for fatigue design.) Thecracks then propagated through the flange and intothe web. After cracking the downstream face of theflanges was 0.5-in. out-of-vertical alignment.

1-2

ETL 1110-2-34630 Sep 93

(2) Quick repairs were performed by operationspersonnel, without input from engineering person-nel. The web crack was filled with weld. Theflange cracks were gouged and welded, then twosmall bars were fillet welded across the crack. Thebar material is unknown. These repairs served toget the gate back into service immediately. How-ever, reliable long-term repairs also should bedeveloped and implemented.

c. Submersible lift gate.

(1) This project has a submersible lift gate asthe main, operational, upstream lock gate. The gateconsists of two leaves with six horizontal girdersspanning 110 ft. Several cracks were discovered inone leaf while the lock was out of service for otherrepairs. Subsequent detailed inspection identifiedover 100 cracks in girder flanges and bracing mem-bers. One crack extended through the downstreamflange of a horizontal girder and 3 ft into the8-ft-deep web.

(2) This gate was subjected to a detailed inves-tigation of the cause of the cracking. The studyidentified several contributing factors: the originaldesign had ignored a loading case and had includedimproper loading assumptions; limit switches wereimproperly stopping the gate before it reached itssupports; the design ignored higher stresses causedby eccentric connections on the downstream face;most of the original welds did not meet currentAmerican Welding Society (AWS) quality stan-dards; the steel for the gate had a low fracturetoughness, ranging from 5 ft-lb at 32oF to 15 ft-lbat 70 oF.

(3) Repair procedures were designed by engi-neering personnel for this gate. However, thespecified weld procedures were not used by thecontractor, and the welders were not properly quali-fied per AWS requirements. These facts may havecaused inadequate repair welds, which duplicatespart of the causes of the original cracking problem.

5. Summary

The preceding examples represent only a few of thesteel cracking problems which have occurred onCorps of Engineers projects. It is evident that steelfatigue and fracture are real problems. Engineering,construction, and operations personnel should beaware of this and of the preventive proceduresneeded to minimize such problems. Prevention isbest accomplished through proper design and con-struction, followed by adequate maintenance andinspection. However, many existing steel structuresmay be susceptible to fatigue and fracture problems.When cracks are discovered, engineering personnelshould evaluate the reliability or remaining life ofthe structure, determine the need for repairs, anddevelop adequate repair plans. When fracturesoccur, operations and engineering personnel shouldwork together to investigate the causes and developreliable repair plans. Enclosures 2 through 6 pre-sent methods for inspection and evaluation of exist-ing steel lock gates. These procedures should befollowed to identify and correct deficiencies beforethey result in serious failures.

1-3

ETL 1110-2-34630 Sep 93

CAUSES OF DETERIORATION

1. Corrosion

Corrosion is degradation of a material by reactionwith its environment. All corrosion processes haveelectrochemical reactions as their bases. Some arepurely electrochemical, such as galvanic, pitting,crevice, and general corrosion, whereas others resultfrom the action of chemical plus mechanical factors,such as erosion and stress corrosion.

a. General corrosion is characterized by anuniform attack over the entire exposed surface withminimal variation in the depth of damage. The rateof attack is usually predictable, and catastrophicfailure does not often result. Galvanic corrosionoccurs when two or more dissimilar metals are incontact and placed in an electrolyte such as water.A potential difference in the metals causes a flowof current between them, and the more active metal(anode) undergoes accelerated corrosion whereascorrosion in the less active metal (cathode) isretarded or eliminated. Galvanic corrosion can beminimized by use of coatings and by keeping theanode large relative to the cathode. Pitting is aform of localized corrosion where the attack isconfined to numerous small cavities on the metalsurface. The length/depth ratio of the pit is usuallyequal to or greater than 1. The pitts can act asstress risers and promote nucleation of fatiguecracks. Failure due to pitting corrosion may berapid and without warning. Crevice corrosion isassociated with confined spaces (< 0.001 in.)formed by close fitting mechanical configurationssuch as tapped joints, washers, and lap joints.

b. Stress corrosion involves the occurrence ofboth chemical and mechanical interactions. Fourbasic requirements are necessary to cause stresscorrosion cracking: a susceptible alloy, an aggres-sive environment, applied or residual tensile stress,and time. The rate of attack is rapid at the cracktip and much less rapid at the sides.

c. The paint system and cathodic protectionsystems should be inspected to assure that protec-tion is being provided against corrosion. The effectof corrosion on the strength, stability, and service-ability of lock gates must be evaluated. The type of

corrosion and cause should be identified to assurethat a thorough evaluation is performed. Ultrasonicequipment and gap gauges are available to measureloss of material. The progressive loss of materialcan increase deflections and result in failure byoverstressing, buckling, or fracture.

2. Unusual Loads

Lock gates are designed to resist loads from selfweight, hydraulic, and boat impact as discussed inEM 1110-2-2703. Dynamic loading due to hydrau-lic flow and impact loading due to vessel collisionis currently unpredictable. The dynamic loadingmay be caused by hydraulic flow at the seals orwhen lock gates are used to supplement chamberfilling or skim ice and debris. Impact loading canoccur from malfunctioning equipment on the vesselor operator error. Furthermore, unusual loadingsmay occur from malfunctioning limit switches ordebris trapped at interfaces between moving parts.In addition, unusual loads may develop on gatessupported by walls that are settling or moving.These unusual loads can cause overstressing andlead to deterioration of the lock gates.

3. Fatigue

a. Most structures are subjected to repeatedcyclic loading. Fatigue is the process of cumulativedamage caused by repeated cyclic loading. Fatiguedamage occurs at stress concentrated regions wherethe localized stress exceeds the yield stress of thematerial. After a certain number of cyclic loads,the accumulated damage causes the initiation andpropagation of a crack.

b. Total fatigue life is the sum of the crackinitiation and the crack propagation to a critical size(Barsom and Rolfe 1987). The main concern infatigue assessment of welded lock gates is to deter-mine the time required for failure to occur. Thepropagation life is governed by the rate of subcriti-cal crack growth. Refer to Enclosure 6 foradditional discussion on fatigue.

Enclosure 2 2-1

ETL 1110-2-34630 Sep 93

4. Fracture

For strength and economic reasons, EM 1110-2-2703 recommends that lock gates be fabricatedusing structural-grade carbon steel. Standards suchas American Society for Testing and Materials(ASTM) A6 or ASTM A898 (1991a,e) have beendeveloped to establish allowable size and number ofdiscontinuities for base metal used to fabricate lockgates. In addition, EM 1110-2-2703 also recom-mends that the gates be welded in accordance withthe Structural Welding Code-Steel (AmericanNational Standards Institute (ANSI)/AWS (1992).This code provides a standard for limiting the sizeand number of various types of discontinuities thatdevelop during welding. Although these criteriaexist, when a lock gate goes into service it doescontain discontinuities.

a. When tensile stresses are applied to a bodythat contains a discontinuity such as a sharp crack,the crack tip tends to open. For cases where plasticdeformation is constrained to a small zone at thecrack tip (plane-strain condition), the fracture insta-bility can be predicted using linear elastic fracturemechanics (LEFM) concepts. The fundamentalprinciple of LEFM is that the stress field ahead of asharp crack in a structural member can be charac-terized in terms of a single parameter,K. K is thestress-intensity factor and has units of kips persquare inch-√in. The stress-intensity factor isrelated to both the nominal stress and the geometryof the existing discontinuity. When the crack isopening with the two fracture surfaces displacedperpendicular to each other in opposite directions,the displacement is referred to as modeI. Thestress-intensity factor during crack opening ormodeI displacement is referred to asKI.

b. Another underlying principle of fracturemechanics is that unstable fracture occurs when thestress-intensity factor at the crack tip reaches a criti-cal value. For modeI displacement and for smallcrack-tip plastic deformation (plane-strain condi-tion), the critical stress-intensity factor for fractureinstability is designatedKIc. The critical stress-intensity factor represents the ability of the materialto withstand a given stress-field intensity at the tipof a crack and to resist tensile crack extension.Thus,KIc represents the fracture toughness of aparticular material and is a function of temperatureand loading rate. When a structural member con-tains a discontinuity, the stress-intensity factor,KI,

should be kept below the critical stress-intensityfactor, KIc, at all times to prevent brittle fracture.

c. Brittle fracture is a sudden catastrophicfailure which occurs suddenly without prior plasticdeformation and can occur at nominal stress levelsbelow yield. Brittle fracture becomes more pre-dominate as member thickness, constraint, andloading rates increase and as temperature decreases.Frequently, plates 1-1/2 in. in thickness and greaterare used as primary welded structural componentson hydraulic gates. It is not uncommon to see suchthick plates used as gate flanges, embeddedanchorage at the top of gates, hinge and operatingequipment connections, diagonal bracing, lifting orjacking assemblies, or platforms to support oper-ating equipment that actuates the gates. In addition,thick castings, such as sector gears used for operat-ing lock gates may be susceptible to brittle fracture.Cracking has been experienced on lock gates duringfabrication and after the thick assemblies are weldedand placed into service.

d. For many structural applications where low-to medium-strength steels are used, the materialthickness is not sufficient to maintain smallcrack-tip plastic deformation under slow loadingconditions at normal service temperatures. Conse-quently, the LEFM approach is invalidated by theformation of large plastic zones and elastic-plasticbehavior in the region near the crack tip. Onemethod frequently used to analyze discontinuitieswhen elastic-plastic conditions exist is the crack-tipopening displacement (CTOD) method (BritishStandards Institution 1980). The LEFM and CTODmethods are discussed in detail in Enclosure 6.

5. Design Deficiencies

Many existing lock gates were designed during theearly and mid-1900’s. Analysis and design technol-ogies have significantly improved to the currentmethodology for gate design. Original design load-ing conditions may no longer be valid for the exist-ing gate operation and overstress conditions mayexist. Current information, such as fatigue andfracture control in structures, was not availablewhen many of the initial designs were performed.Consequently, low category fatigue details and lowtoughness materials exist on some lock gates. Inaddition, the amount of corrosion anticipated in theoriginal design may not accurately reflect actual

2-2

ETL 1110-2-34630 Sep 93

conditions, and structural members may now beundersized. To properly evaluate existing lockgates, it is important that the analysis and designinformation for the gate be reviewed to assure nodesign deficiencies exist.

6. Fabrication Discontinuities

Welded fabrication can contain various types ofdiscontinuities. Discontinuities in regions near theweld are of special concern, since high-tensileresidual stresses develop from the welding process.There are two reasons that fabrication discontinu-ities reduce the strength of welded gates. First, thepresence of the discontinuities decreases the sec-tional areas, and second, stress becomes concen-trated around the discontinuities. The effect ofweld discontinuities on structural strength dependsupon the nature and size of discontinuities, type ofmaterial, and type of loading. Discontinuities thatexist during initial fabrication are rejectable onlywhen they exceed specified requirements in termsof type, size, distribution, or location as specifiedby ANSI/AWS (1992). In addition, industry stan-dards have improved in the area of materialprocessing and fabrication. Therefore, existingstructures may have included joint preparation and

welds which may not be acceptable according tocurrent standards.

7. Operation and Maintenance

Proper operation and maintenance of lock gates isnecessary to prevent structural deterioration. Ifmoving connections are not lubricated properly, thebushings will wear and result in misalignment ofthe gate. The misalignment will subsequently wearcontact blocks and seals, and unforseen loads maydevelop. Overstressing and vibrational loads couldthen develop and reduce the life of the gate. Mal-functioning limit switches and debris along the gatepath can also induce detrimental loads and wear.As discussed in this enclosure, paragraph 1, it isessential that an effective coating system be main-tained on the gates to minimize corrosion. Further-more, when cathodic protection is necessary, it, too,must be properly operated and maintained. In addi-tion, to assure that necessary torsional stability isprovided during opening and closing of miter gates,it is essential that the prestress in the diagonals bemaintained. In addition, proper maintenance oftimber fenders and bumpers is necessary to provideprotection to the gate and minimize deterioration.

2-3

ETL 1110-2-34630 Sep 93

PERIODIC INSPECTION

1. Purpose of Inspection

Existing welded lock gates are subjected to condi-tions which could cause structural deterioration andpremature failure. The causes of deterioration arediscussed in Enclosure 2. To assure prematurefailures are averted and identify future maintenancerequirements, periodic inspections are performed asdiscussed in Engineer Regulation (ER) 1110-2-100,"Periodic Inspection and Continuing Evaluation ofCompleted Civil Works Structures." Periodicinspections on lock gates are primarily visualinspections. If the periodic inspection indicates thata gate may be distressed, a more detailed inspectionand evaluation may be necessary. This detailedinspection may require nondestructive and/ordestructive testing as discussed in Enclosures 4and 5. The information obtained from the inspec-tions and tests will then be used to perform a struc-tural evaluation as discussed in Enclosure 6 andmake a recommendation for future action. Thisenclosure will further discuss the visual inspectionwhich should be performed during the periodicinspection.

2. Inspection Procedures

The periodic inspection procedure should includethe following steps:

a. Review documentation on gate design,operational history, and maintenance record.

b. Identify critical members and connections.

c. Develop plan for visual inspection.

d. Inspect for weld condition and surfacediscontinuities.

e. Inspect for corrosion conditions.

f. Observe gate operation (and cathodic protec-tion, if applicable).

g. Document weld, discontinuity, and corrosionconditions.

h. Conduct initial evaluation.

3. Critical Members and Connections

a. The periodic inspection should assure thatall critical members and connections are fit forservice until the next scheduled inspection. Criticalmembers and connections are those structural ele-ments whose failure would render the gate inoper-able. Fitness for service means that the materialand fabrication quality are at an appropriate levelconsidering risks and consequences of failure(Enclosure 6).

b. Critical gate members and connections canbe determined from structural analysis of the gate.This should include local stress concentrations andfatigue considerations. In addition, effects fromexisting corrosion and reduced weld quality orassociated residual stresses should be considered.This analysis will require information pertaining tothe existing mechanical properties of the structuralmaterial and weld (i.e. strength, toughness, ductility)and the location, type, size, and orientation of anyknown discontinuities.

4. Visual Inspection

a. The inspector should look closely at themembers and connections and not just view themfrom the top of the lock wall. Visual inspectionsshould be performed with an emphasis on criticalgate members and connections as discussed in para-graph 3 of this enclosure. Historically, distressedgate members and connections have been located inareas subject to high structural loads or stressranges, geometric stress concentrations, corrosion-promoting conditions, and thick plates.

b. Inspectors should use various measuringscales and weld gauges for checking the dimensionsof the weld bead. Boroscopes, flashlights, andmirrors may be necessary to inspect areas of limitedaccessibility. Hand-tools may be necessary forcleaning the surface for inspection.

Enclosure 3 3-1

ETL 1110-2-34630 Sep 93

5. Other Inspection Methods

Inspection methods other than visual inspection maybe used for the periodic inspection of lock gates, ifnecessary. These methods may include penetrant,magnetic particle, ultrasonic, and eddy-currentinspections. These inspection methods are dis-cussed in Enclosure 4.

6. Initial Evaluation

The most common problems identified by a visualinspection are discovery of weld bead noncompli-ance, with respect to the ANSI/AWS D1.1-92(1992), Structural Welding Code-Steel, surfacecracks, fracture of structural members, and deterio-ration from corrosion. For weld bead noncompli-ances, the initial evaluation will be based onchecking with the ANSI/AWS D1.1-92 code accep-tance criteria. If surface cracks or fractured mem-bers are discovered during the periodic inspections,detailed inspection and evaluation shall be per-formed for the entire gate. The strength and stabil-ity of corroded members should be calculated. Loss

of material due to corrosion can often be deter-mined using ultrasonic inspection methods. If thestrength or stability under the existing conditionsdoes not meet the design criteria, then the loadsmust be reduced by modifying the operational pro-cedures or the section should be replaced or rebuilt.

7. Inspection Intervals

The maximum time interval between periodicinspections of lock gates is established in ER 1110-2-100. Visual inspections should also be performedif unusual loading situations occur. Such situationsinclude barge impact, earthquake, excessive iceload, frictional forces increase between seals andembedded plates, and movement of the supportingmonoliths. Additional detailed inspections may berequired to pursue concerns developed from theperiodic inspections or investigate reported distressfrom lock personnel. If discontinuities exist, frac-ture mechanics concepts can also be applied todetermine appropriate inspection intervals as dis-cussed in Enclosure 6.

3-2

ETL 1110-2-34630 Sep 93

DETAILED INSPECTION

1. Purpose of Inspection

a. If distressed gate members or connectionsare identified in the periodic inspection or deteriora-tion in structural performance is assessed from theinitial evaluation, then the entire gate should receivea more detailed inspection of the distressed mem-bers, and connections should be evaluated. Thisenclosure presents a summary of various inspectionmethods, guidance in selecting inspection methods,inspector qualifications, code acceptance criteria,and applicable source documents that may aid inperforming a detailed inspection.

b. Detailed inspections may be also used aspart of a damage-tolerance fracture control planwhich has been used to optimize the use of weldedstructures in many industries. This fracture controlconcept is based on the fact that presence of crack-like discontinuities in the structural members orconnections does not necessarily mean the end ofthe structure’s service life. An integrated approachusing scheduled inspections on the flawed membersand analysis of fracture/fatigue resistance of thesame members can maintain satisfactory structuralperformance. The cost for repair or replacement ofthe flawed members can therefore be balancedagainst the inspection cost.

c. To develop schedules for inspection whenthe damage-tolerance fracture control plan is used,fracture mechanics theories must be applied. Theinspection periods can be determined by fatiguepropagation analysis of the cracked structural mem-bers. The crack growth history from a detectablesize to the critical size can be predicted using thepropagation laws (e.g. Paris’s crack growth law).Time interval between inspections should be afraction of this crack growth life. The optimumnondestructive testing (NDT) intervals vary withservice conditions and the discontinuity conditions.These inspection intervals should be short enoughthat the nondetectable cracks at the precedinginspections do not have time to propagate to failurebefore the next scheduled inspection.

d. A procedure for planning the inspectionschedules from the crack growth analysis ispresented in Enclosure 6.

2. Selecting Inspection Methods

a. NDT methods are essential for field inspec-tion of existing lock gates. NDT can be used toimprove structural reliability by detecting discon-tinuities for appropriate repair. NDT methods differfrom destructive testing methods which damage orimpair the serviceability of the items tested.

b. The six NDT methods commonly used intoday’s industries are visual (VT), penetrant (PT),magnetic-particle (MT), radiographic (RT), ultra-sonic (UT), and eddy-current (ET). Selection of anNDT method for inspection depends on a number ofvariables, including the nature of the discontinuity,accessibility, joint type and geometry, material type,detectability and reliability of the inspectionmethod, inspector qualifications, and economicconsiderations. A general guide for selecting NDTmethods for field inspection is given in Table 4-1,this enclosure.

3. Inspector Qualifications

For an inspection to be worth performing, theinspector must be qualified. Corps personnel areoften not adequately trained in inspection methods;therefore, inspections are often performed via con-tract with inspection specialists. The followingqualification requirements apply to all inspectors,whether government or contractor employees.

a. Qualification in NDT methods.

(1) The effectiveness of NDT depends on thecapabilities of the person who performs the test.Inspectors performing NDT should be qualified inaccordance with the American Society for Non-destructive Testing (ASNT) Recommended PracticeNo. SNT-TC-1A (ASNT 1980). The SNT-TC-1Adocument is a guide to establish practices for train-ing, qualification, and certification of NDT person-nel. Three basic levels of qualification are definedin SNT-TC-1A as follows:

(a) NDT Level I: An NDT Level I individualshall be qualified to properly perform specific cali-brations, specific NDT, and specific evaluations for

Enclosure 4 4-1

ETL 1110-2-34630 Sep 93

Table 4-1Selection Guide for Inspection Method

Method Applications Advantages Disadvantages

Visual Surface discontinuities Economical, fast. Limited to visual acuity of the inspector.

Liquid Surface cracks and porosity Relatively inexpensive and Cleaning is needed before and afterpenetrant reasonably rapid. inspection. Surface films hide defects.

Magnetic Surface discontinuities and Relatively economical and Applicable only to ferromagneticparticle large subsurface voids expedient. materials.

Radiographic Voluminous discontinuities Provides a permanent Planer discontinuities must be favorablySurface and internal record. aligned with radiation beam. Cost of

discontinuities equipment is high.

Ultrasonic Most discontinuities Sensitive to planer type Small, thick parts may be difficult todiscontinuities. High inspect. Requires a skilledpenetration capability. operator.

Eddy current Surface and subsurface Painted or coated surfaces Many variables can affect the testdiscontinuities can be inspected. signal.

High speed.

acceptance or rejection determinations according towritten instructions and to record results.

(b) NDT Level II: An NDT Level II individualshall be qualified to set up and calibrate equipmentand to interpret and evaluate results with respect toapplicable codes, standards, and specifications. TheNDT Level II individual shall be able to organizeand report the results of NDT.

(c) NDT Level III: An NDT Level III indivi-dual shall be capable of establishing techniques andprocedures; interpreting codes, standards, and proce-dures; and designating the particular NDT methods,techniques, and procedures to be used.

(2) Certification of all levels of NDT personnelis the responsibility of the employer. The employermust establish a written practice for the control andadministration of NDT personnel training, examina-tion, and certification.

b. Qualification in weld inspection.

(1) Welding inspectors are responsible forjudging the quality of the product in relation tosome form of written specification. The followingqualifications are necessary for individuals to ade-quately inspect welds:

(a) A welding inspector must be familiar withengineering drawings and able to interpretspecifications.

(b) A welding inspector should be familiarwith welding processes and welding procedures.

(c) A welding inspector should be able tomaintain adequate records.

(d) A welding inspector should have passedan eye examination with or without correctivelenses to prove:

• Near vision acuity of Snellen English, orequivalent, at 12 in.

• Far vision acuity of 20/40, or better.

(2) In addition, one of the following threerequirements is necessary to qualify an individual asa weld inspector for a lock gate:

(a) Current or previous certification as anAWS Certified Welding Inspector (CWI) in accor-dance with the provisions of AWS QC1-88, Stan-dard and Guide for Qualification and Certificationof Welding Inspectors (ANSI/AWS 1988).

4-2

ETL 1110-2-34630 Sep 93

(b) Current or previous qualification by theCanadian Welding Bureau (CWB) to the require-ments of the Canadian Standard Association (CSA)Standard W178.2, Certification of WeldingInspectors (CSA 1917).

(c) An engineer or technician who, by training,experience, or both, in metals fabrication, inspectionand testing, is competent to perform inspection ofthe work.

4. Inspection Reporting

A report should be completed by the inspector atthe time of inspection. It should show the location,size, orientation, and classification of each disconti-nuity. The following information should be identi-fied and recorded in the report:

a. Identification and location of inspectedstructures.

b. Date and time of inspection.

c. Type of inspection.

d. Inspection procedure.

e. Inspection system (equipment).

f. Inspector identity and level.

g. Record of discontinuities detected.

5. Summary of NDT Methods

a. Detailed visual inspection (VT).

Detailed VT inspection uses the same inspectiontools and procedure as that described inEnclosure 3, except that, with a knowledge ofexisting discontinuities in a structural member orconnection from periodic inspections, a more con-centrated examination is performed. The type,geometry, size, location, and orientation of thediscontinuities must be quantitatively determined.The entire structure may be inspected rather thanjust representative members or connections.

(1) Advantages. VT inspection is useful forchecking the presence of surface discontinuities. It

is simple, quick, and easy to apply. It requires nospecial equipment other than good eyesight, some-times assisted by simple and inexpensiveequipment.

(2) Disadvantages and limitations. A majordisadvantage of VT inspection is the need for aninspector who has considerable experience andknowledge in many different areas. Although VTinspection is an invaluable method for detectingsurface discontinuities, it is less reliable in detectingand quantifying small surface discontinuities ordetecting subsurface discontinuities.

(3) Applicable document. Material pertainingto VT inspection is included in ANSI/AWSB1.10-86, "Guide for the Nondestructive Inspectionof Welds" (ANSI/AWS 1986).

b. Penetrant inspection (PT).

PT inspection is also a method used to detect andlocate surface discontinuities. Liquid penetrants canseep into various types of minute surface openingsby capillary action. Therefore, this process is wellsuited for detecting discontinuities such as surfacecracks, overlaps, porosity, and laminations. PTinspection can be performed using visible dye orfluorescent dye visible with ultraviolet light. Threedifferent penetrants commonly used with either dyeare water washable, solvent removable, and postemulsifiable. The various penetrant inspectionsystems are listed in an order of decreasing inspec-tion sensitivity and operational cost as follows:

• Post emulsifiable fluorescent dye

• Solvent removable fluorescent dye

• Water washable fluorescent dye

• Post emulsifiable visible dye

• Solvent removable visible dye

• Water washable visible dye

(1) Advantages. PT inspection is relativelyinexpensive and reasonably rapid. Equipment

4-3

ETL 1110-2-34630 Sep 93

generally is simpler and less costly than that formost other NDT methods.

(2) Disadvantages and limitations. The majorlimitation of PT inspection is that it can detect onlydiscontinuities that are open to the surface. Anotherdisadvantage is that the surface roughness of theobject being inspected may affect the PT inspectionresults. Extremely rough or porous surfaces mayproduce false indications. Some substances in thepenetrants can affect structural materials. If pene-trants are corrosive to the gate material, they shouldbe avoided.

(3) Applicable documents.

(a) ASTM E165-91: Standard Test Method forLiquid Penetrant Examination (ASTM 1991h).

(b) ASTM E1316-92: Standard Terminologyfor Nondestructive Examinations (ASTM 1992f).

(c) AWS B1.10-86: Guide for the Nondestruc-tive Inspection of Welds (ANSI/AWS 1986).

c. Magnetic particle inspection (MT).

MT inspection is used to detect surface or near-surface discontinuities in ferromagnetic materials.Magnetic fields can be generated by yokes, coils,central conductors, prod contacts, and induced cur-rent. When the material is magnetized, magneticdiscontinuities that lie in a direction generally trans-verse to the direction of the magnetic field willcause a leakage field at the surface of the material.The presence of this leakage field is detected by theuse of fine ferromagnetic particles applied over thesurface, some of the particles being gathered andheld by the leakage field. This collection of parti-cles indicates the discontinuities. Several magneticparticle materials commonly used for MT inspectionare dry powders (i.e. suitable for field inspection oflarge object), wet magnetic particles suspended inwater or light oil (i.e. suitable for very fine or shal-low discontinuities), magnetic slurry suspended inheavy oil, and magnetic particles dispersed in theliquid polymers to form solid indications.

(1) Advantages. The MT inspection is a sensi-tive means of detecting small and shallow surfaceor near-surface discontinuities in ferromagneticmaterials. The cost of MT inspection is consider-ably less expensive than radiographic or ultrasonic

inspection. MT inspection is generally faster andmore economical than penetrant inspection. Com-pared to PT inspection, MT inspection has theadvantage of revealing cracks filled with foreignmaterial.

(2) Disadvantages and limitations. MTinspection is limited to ferromagnetic material. Forgood results, the magnetic field must be in a direc-tion that will intercept the direction of the discon-tinuity. Large currents sometimes are required forvery large parts. Care is necessary to avoid localheating and burning of surfaces at the points ofelectrical contact. Demagnetization is sometimesnecessary after inspection. Discontinuities must beopen to the surface or must be in the near subsur-face to create flux leakage of sufficient strength toaccumulate magnetic particles. If a discontinuity isoriented parallel to the lines of force, it will beessentially undetectable.

(3) Applicable documents.

(a) ASTM E1316-92: Standard Terminologyfor Nondestructive Examinations (ASTM 1992f).

(b) ASTM E709-91: Standard Guide forMagnetic Particle Examination (ASTM 1991l).

(c) ANSI/AWS B1.10-86 (ANSI/AWS 1986):Guide for the Nondestructive Inspection of Welds.

d. Radiographic inspection (RT).

RT inspection is based on differential absorption ofpenetrating radiation by the material beinginspected. Radiation from the source is absorbedby the test piece as the radiation passes through it.The discontinuity and its surrounding materialabsorb different amounts of penetrating radiation.Thus, the amount of radiation that impinges on thefilm in the area beneath the discontinuity is differ-ent from the amount that impinges in the adjacentarea. This produces a latent image on the film.When the film is developed, the discontinuity canbe seen as a shadow of different photographic den-sity from that of the image of the surroundingmaterial. Evaluation of the radiograph is based ona comparison of these differences in photographicdensity. The dark regions represent the more easilypenetrated parts (i.e. thin sections and most types ofdiscontinuities) while the lighter regions representthe more difficult areas to penetrate (i.e. thick

4-4

ETL 1110-2-34630 Sep 93

sections). An essential element to the radiographicprocess is film, a thin transparent plastic basecoated with fine crystals of silver bromide(emulsion).

(1) Advantages.

(a) RT inspection has an ability to detect sur-face and internal discontinuities.

(b) It is generally not restricted by the type ofmaterial or grain structure.

(c) It provides a permanent record for futurereview.

(2) Disadvantages and limitations.

(a) Discontinuities must be favorably alignedwith the radiation beam for reliable detection.

(b) It presents a potential radiation hazard topersonnel.

(c) The cost of radiographic equipment, facili-ties, and safety programs is relatively high.

(d) Accessibility to both sides of the parts to beinspected is required.

(e) It is difficult to apply for field inspections.

(f) It is a time consuming process compared toother NDT methods.

(3) Applicable documents.

(a) ASTM E 94-91: Standard Guide forRadiographic Testing (ASTM 1991g).

(b) ASTM E142-92: Standard Method forControlling Quality of Radiographic Testing (ASTM1992c).

(c) ASTM E242-91: Standard ReferenceRadiographs for Appearances of RadiographicImages as Certain Parameters are Changed (ASTM1991k).

(d) ASTM E1316-92: Standard Terminologyfor Nondestructive Examination (ASTM 1992f).

(e) ASTM E747-90: Standard Test Methodfor Controlling Quality of Radiographic Examina-tion Using Wire Penetrameters (ASTM 1990h).

(f) ASTM E999-90: Standard Guide for Con-trolling the Quality of Industrial Radiographic FilmProcessing (ASTM 1990i).

(g) ASTM E1025-84: Standard Practice forHole-Type Image Quality Indicators Used forRadiography (ASTM 1989c).

(h) ASTM E1032-92: Standard Method forRadiographic Examination of Weldments (ASTM1992e).

(i) ANSI/AWS B1.10-86: Guide for the Non-destructive Inspection of Welds (ANSI/AWS 1986).

(j) ANSI/AWS D1.1-92: Structural WeldingCode-Steel (Chapter 6: Inspection) (ANSI/AWS1992).

e. Ultrasonic inspection (UT).

UT inspection is a nondestructive method whichuses high-frequency sound waves to detect surfaceand internal discontinuities. The sound wavestravel through the materials to be inspected and arereflected from surfaces refracted at a boundarybetween two substances and diffracted at edges oraround obstacles. The reflected sound beam isdetected and analyzed to define the presence andlocation of discontinuities. Cracks, laminations,shrinkage cavities, pores, and other discontinuitiesthat act as metal-gas interfaces can be easilydetected. Inclusions and other inhomogeneities inthe metal can also be detected. All surfaces of thepart to be examined should be free of weld spatter,dirt, grease, oil, paint, and loose scale. UT inspec-tion is usually performed with longitudinal waves orshear waves (i.e. angle beam). Most UT inspec-tions for discontinuities are performed using angle-beam technique. The pulse-echo method withA-scan is most commonly used for inspection ofwelds. The most commonly used frequencies arebetween 1 and 5 MHz, with sound beams at anglesof 0, 45, 60, and 70 deg.

(1) Advantages.

4-5

ETL 1110-2-34630 Sep 93

(a) Superior penetrating power allows thedetection of discontinuities deep in the part.

(b) High sensitivity permits the detection ofsmall discontinuities.

(c) Great accuracy in determining the size,position, and the shape of discontinuities.

(d) Almost instantaneous indications of dis-continuities provided.

(e) Ultrasonic inspection is not hazardous topersonnel and has no effect on materials.

(2) Disadvantages and limitations.

(a) Manual operation requires careful attentionby experienced technicians.

(b) Parts that are rough, irregular in shape, verysmall, or inhomogeneous are difficult to inspect.

(c) Reference standards are needed for cali-brating the equipment and for characterizingdiscontinuities.

(d) Interpretation requires experiencedtechnicians.

(3) Applicable documents.

(a) ASTM A435/A435M-90: Standard Speci-fication for Straight-Beam Ultrasonic Examinationof Steel Plates (ASTM 1990a).

(b) ASTM A577/A577M-90: Standard Speci-fication for Ultrasonic Angle-Beam Examination ofSteel Plates (ASTM 1990c).

(c) ASTM E114-90: Standard Practice forUltrasonic Pulse-Echo Straight-Beam Examinationby the Contact Method (ASTM 1990d).

(d) ASTM E164-90: Standard Practice forUltrasonic Contact Examination of Weldments(ASTM 1990e).

(e) ASTM E214-68: Standard Practice forImmersed Ultrasonic Examination by the ReflectionMethod Using Pulsed Longitudinal Waves (ASTM1991j).

(f) ASTM E1316-92: Standard Terminologyfor Ultrasonic Examination (ASTM 1992f).

(g) AWS B1.10-86: Guide for the Nonde-structive Inspection of Welds (ANSI/AWS 1986).

(h) ANSI/AWS D1.1-92: Structural WeldingCode-Steel (Chapter 6: Inspection) (ANSI/AWS1992).

f. Eddy-current inspection (ET).

ET inspection is an electromagnetic NDT methodwhich is based on the principles of electromagneticinduction. When an alternating current is passedthrough a coil, eddy current is created in the mater-ial being tested by an alternating magnetic field.The test coil is electronically monitored to detectthe changes of magnetic field caused by the inter-action between the eddy currents and the initialfield. Any changes in the eddy currents due toinhomogeneities in the material are detected; there-fore, any surface or subsurface discontinuities thatappreciably alter the normal flow of eddy currentscan be detected by ET inspection. Because ETinspection is an electromagnetic induction tech-nique, it does not require direct contact betweenprobe and the material being tested. The method isbased on indirect measurement, and the correlationbetween the instrument readings and the structuralcharacteristics of the material being inspected mustbe carefully established.

(1) Advantages.

(a) Since direct contact between probe and thematerial is not required, painted, or coated, mater-ials can be inspected.

(b) ET inspection is adaptable to high-speedinspection.

(2) Disadvantages and limitations.

(a) The test material must be an electricalconductor.

(b) Some internal discontinuities cannot bedetected by eddy-current inspection.

4-6

ETL 1110-2-34630 Sep 93

(c) Since many variables can affect an eddy-current signal, variables of no concern must beseparated from those of interest.

(3) Applicable documents.

(a) ASTM E1316-92: Standard Terminologyfor Nondestructive Examination (ASTM 1992f).

(b) ANSI/AWS B1.10-86: Guide for the Non-destructive Inspection of Welds (ANSI/AWS 1986).

6. Acceptance Criteria for NDT Results

a. The common weld discontinuities detectedfrom various NDT methods can be classified intoplanar and nonplanar types. Planar type discontinu-ities include cracks, delaminations or laminar tear-ing, and sometimes incomplete joint penetration orincomplete fusion. The nonplanar type discontinu-ities are volumetric weld discontinuities whichinclude porosity, slag or tungsten inclusions, under-cut, underfill, and overlap. Figure 4.1 shows thesecommon types of weld discontinuities defined byANSI/AWS B1.10-86, Guide for the NondestructiveInspection of Welds (ANSI/AWS 1986).

b. The results obtained from various NDTinspections are usually assessed according to thecode acceptance criteria. The recommended accep-tance criteria for weld discontinuities are presentedin the ANSI/AWS D1.1-92 Structural WeldingCode (ANSI/AWS 1992). Repair or replacement ofstructural members or connections which containunacceptable discontinuities (i.e. flaws) may berequired. However, fracture mechanics analysismay be conducted to reassess these unacceptablediscontinuities. A maintenance schedule may bedeveloped in lieu of immediate repair or replace-ment of the distressed members or connectionsusing the damage-tolerance fracture control plan(Enclosure 6).

c. The ANSI/AWS D1.1-92 Structural WeldingCode acceptance criteria for various NDT inspectionresults can be summarized in three perspectives:weld profile, static loading case, and dynamic

loading case. Weld profile is a code compliance for

Figure 4-1. Weld discontinuities (ANSI/AWS1986; copyright permission granted by AmericanWelding Society)

weld quality. Inspection for this code compliance isusually made by visual inspection with the aid of aweld gauge. The purpose of this code complianceis to provide information on the structural fitness ofthe welds. However, weld profile noncompliancemay be acceptable if an engineering assessment isconducted.

d. The code acceptance criteria recognize theeffect of dynamic loading on the structures asopposed to the statically loaded case. Planar typediscontinuities are not acceptable in either case.Permissible conditions on nonplanar type disconti-nuities are specified in the code criteria with smallerallowances for the dynamically loaded structures.Engineering analyses may be conducted to assessthe structural significance of the unacceptablediscontinuities in both instances. A damage-tolerance fracture control plan may be used ratherthan repair or replacement of the distressedmembers or connections.

4-7

ETL 1110-2-34630 Sep 93

MATERIAL AND WELD TESTING

1. Purpose of Testing

a. Distressed gate members and connectionsidentified from NDT inspection may continue toperform the structural functions with adjustments inload conditions or under a reduced safety factorwithout load adjustment. Engineering assessmentsshould include fracture and fatigue analysis asdiscussed in Enclosure 6. Mechanical properties ofthe structural members and welds are usuallyneeded in the analysis.

b. For lock gates fabricated in recent years, thematerials used for the structural members and weldsare usually well documented and can be identifiedfrom the design drawings. For older gates, how-ever, information on mechanical properties of thestructural materials or welds may not be readilyavailable. Mechanical tests of these materials andwelds are sometimes required to determine neces-sary information for fracture and fatigue analyses.In addition, it may be required to determine thechemical composition of unknown materials toassist in selecting the appropriate NDT inspectionmethod, performing corrosion assessment, and con-ducting fracture evaluation.

2. Selection of Samples from ExistingStructure

Material information that is frequently required tostructurally evaluate a welded gate includes chemi-cal composition, tensile strength, bend ductility,fillet weld shear strength, hardness, and fracturetoughness. The test samples may be taken from thematerials left from original fabrication, removedfrom existing gate members or connections, orobtained from weldments made of similar materialswith welding procedures similar to those used in theoriginal fabrication.

3. Chemical Analysis

When the chemical composition of an existing gatematerial is not available, it may be necessary toperform a chemical analysis. This is an importantinitial task in the overall material and weld testingprogram. The information from this analysis will

provide a basis of similarity to other known struc-tural materials for characterizing the properties ofthe unknown gate materials. This information canbe used to assist in selecting appropriate NDTmethods, assessing corrosion problems, conductingfracture analyses, and assessing material weldabilityfor possible repair. A chemical analysis for mate-rial compositions should be in conformance withASTM E30-89 and E350-90 (1989b, 1990f).

4. Tension Test

a. Tension tests on the base metal and weldmetal can provide information on the strength andductility of materials under uniaxial tensile stress.Transverse rectangular tension tests of weld samplesshow the effect of material inhomogeneity and weldquality on the test results. The pertinent dataobtained from a tension test are ultimate tensilestrength, yield strength, Young’s Modulus, percentelongation, percent reduction of cross-sectional area,stress-strain curve, and location and mode of finalfracture.

b. The transverse rectangular tension speci-mens are machined from a butt welded plate, withthe weld crossing in the midsection of the specimen(AWS B4.0-85 (AWS 1985), Figure C-2). Whenweldment thickness is beyond the capacity of testequipment, the weldment is divided through itsthickness into as many specimens as required tocover the full weld thickness. The results of thepartial thickness specimens are averaged to deter-mine the properties of the full thickness joint.

c. Excessively deep machine cuts that willcause specimen bending during testing or that leavetears in the surface of the finished dimensionsshould be avoided. Imperfections present in theguage length, which are incidental to welding,should not be removed.

d. The base metal and weld metal tests areperformed on a tensile testing machine in accor-dance with the requirements of ASTM E8-91(1991f). The machine should be calibrated inaccordance with ASTM E4-89 (1989a). Thetesting procedure is as specified in ASTM E8-91(1991f). The rate of straining should be

Enclosure 5 5-1

ETL 1110-2-34630 Sep 93

between 0.05 and 0.5 in. per inch of guage length,per minute.

e. Material properties are calculated as follows:

(1) Ultimate tensile strength = maximumload/original cross-sectional area in the guagelength.

(2) Yield strength = load at 0.2% offset/originalcross-sectional area in the guage length.

(3) Percent elongation = (Final guage length -original guage length)/original guage length × 100.

(4) Reduction of area: Fit the ends of the frac-tured specimen together and measure the thicknessand width at the minimum cross section. Calculatethe reduced area.

At least two specimens should be tested for eachsample type. The result of the tension test is theaverage of the results of the specimens.

f. Applicable documents.

(1) ANSI B46.1-85: Surface Texture (ANSI1985).

(2) ASTM E4-89: Standard Practices for LoadVerification of Testing Machines (ASTM 1989a).

(3) ASTM E8-91: Standard Test Methods forTension Testing of Metallic Materials (ASTM1991f).

(4) AWS A2.4-86: Standard Symbols forWelding, Brazing, and Nondestructive Examination(AWS 1986).

(5) AWS A3.0-89: Standard Welding Termsand Definitions (AWS 1989).

(6) AWS B4.0-85 - Part C: Tension Testing ofWelded Joints (AWS 1985).

5. Bend Test

a. Guided bend tests are used to evaluate theductility and soundness of welded joints and to

determine incomplete fusion, cracking, delamina-tion, effect of bead configuration, and macrodefectsof welded joints. The quality of welds can be eval-uated as a function of ductility to resist crackingduring bending. The top and bottom surfaces of awelded plate are designated as the face and rootsurfaces, respectively. Face bends have the weldface on the tension side of the bent specimen, andthe weld root is on the tension side for root bends.For thick plates, transverse slices are cut from thewelded joint, and one of the cut side surfacesbecomes the tension side of the bent specimen.

b. When the plate thickness is less than orequal to 3/8 in., two specimens are tested for facebend and two specimens are tested for root bend.When the thickness of the plate is greater than3/8 in., four specimens are tested for side bend.

c. Transverse side bend test specimens (AWSB4.0-85 (AWS 1985), Figure A-5) are used forplates that are too thick for face bend or root bendspecimen. The weld is perpendicular to the longitu-dinal axis of the specimen. The side showing moresignificant discontinuities should be the tension sur-face of the specimen.

d. For a transverse face bend specimen (AWSB4.0-85 (AWS 1985), Figure A-6a), weld is perpen-dicular to the longitudinal axis of the specimen. Theweld face becomes the tension surface of the speci-men during bending. For transverse root bendspecimen (AWS B4.0-85, Figure A-6b), weld isperpendicular to the longitudinal axis of the speci-men. The root surface of the weld becomes thetension surface of the specimen during bending.For all types of bend tests, face, root, and side, thespecimen is tested at ambient temperature, anddeformation should occur in a time period between1/2 and 2 min.

e. During the test, the convex surface of thebent specimen should be examined frequently forcracks or other open defects. If a crack or opendefect is present after bending, exceeding a speci-fied size measured in any direction, the specimen isconsidered to be failed (AWS B4.0-85 (AWS1985)). Cracks occurring on the corners of thespecimen during testing are not considered to fail aspecimen unless they exceed a specified size orshow evidence of defects (AWS B4.0-85).

5-2

ETL 1110-2-34630 Sep 93

f. Applicable documents.

(1) ANSI B46.1-85: Surface Texture (ANSI1985).

(2) ASTM E190-92: Standard Test Method forGuided Bend Test for Ductility of Welds (ASTM1992d).

(3) AWS A2.4-86: Standard Symbols forWelding, Brazing, and Nondestructive Examination(AWS 1986).

(4) AWS A3.0-89: Standard Welding Termsand Definitions (AWS 1989).

(5) AWS B4.0-85 - Part A: Bend Testing ofWelded Joints (AWS 1985).

6. Fillet Weld Shear Test

a. The fillet weld shear test is used to deter-mine the shear strength of fillet welds. The testspecimens are usually made from a weld samplewith welding procedures similar to that used in theoriginal fabrication. During testing, a tensile load isplaced on the specimen to shear the fillet welds.The shear strength of the weld is reported as loadper unit weld length.

b. For longitudinal shear strength, the specimenis prepared in accordance with AWS B4.0-85 (AWS1985), Figure E-1. For transverse shear strength,the test specimen is prepared in accordance withAWS B4.0-85, Figure E-2. The surface contourand size of the fillet welds should be in accordancewith the applicable code or standards.

c. The test is performed on a tensile machine inaccordance with the requirements of ASTM E8-91(ASTM 1991f). The machine should be calibratedin accordance with ASTM E4-89 (ASTM 1989a).The specimen is positioned in the testing machineso that the tensile load is applied parallel to thelongitudinal axis of the specimen. The length,average throat dimension, and legs of each weldshould be measured and reported. The welds aresheared under tensile loads and the maximum ten-sile loads are reported.

d. Shear strength in pounds per square inch iscalculated by dividing the maximum load by the

effective weld throat area (i.e. theoretical throatthickness times total length of fillet weld sheared).At least two specimens are tested. The result of theshear test is the average of the results of the speci-mens. A test is considered invalid if the failure iscaused by a base metal defect. The fracture loca-tion must also be included in the report.

e. Applicable documents.

(1) ASTM E4-89: Standard Practices forLoad Verification of Testing Machines (ASTM1989a).

(2) ASTM E8-91: Standard Test Methods forTension Testing of Metallic Materials (ASTM1991f).

(3) AWS A2.4-86: Standard Symbols forWelding and Nondestructive Testing (AWS 1986).

(4) AWS A3.0-89: Standard Welding Termsand Definitions (AWS 1989).

(5) AWS B4.0-85 - Part E: Fillet Weld ShearTest (AWS 1985).

7. Hardness Test

a. Hardness tests are used in weld evaluationsto provide information on the generic weld proper-ties. Hardness measurements provide indications ofmetallurgical changes caused by welding, metallurg-ical variations and abrupt microstructural discon-tinuities in weld joints, brittleness, and relativesensitivity to cracking under structural loads.

b. Specimens for hardness testing includeas-welded partial or complete assemblies, weld-ments from which the reinforcement has beenremoved, and weld joint cross sections. For hard-ness tests of existing lock gates, the weld reinforce-ment may or may not be removed. When it isremoved, a local area of the reinforcement isground smooth before testing. For large assemblies,portable hardness testers are available that can betransported for use in the field. Microhardnesstesting of weld is usually performed on ground,polished, or polished and etched transverse crosssections of the weld joints.

5-3

ETL 1110-2-34630 Sep 93

c. Hardness testing methods include Brinell,Rockwell, Vickers, and Knoop tests. Selection oftest method depends on hardness or strength of thematerial, the size of the welded joints, and the typeof information desired. The Brinell test produces alarge indentation and is suited for large welds inheavy plates, which is suitable for field evaluations.The Rockwell test produces much smaller indenta-tions than the Brinell test and is more suited forhardness traverses. The Rockwell hardness test isalso suitable for field inspection if a portable testeris used. The Vickers and Knoop tests make rela-tively small indentations and are suited for hardnessmeasurements of the various regions in the weldheat-affected zone and for fine-scale traverses. TheBrinell and Rockwell tests are generally used forhardness measurements of fusion-welded joints inlaboratory condition or field environment.

d. The Brinell hardness test is performed inaccordance with the requirements of ASTM E10-84(ASTM 1984). It is an indentation hardness testusing calibrated machines to force a hard ball intothe surface of the material and to measure the diam-eter of the resulting impression after removal of theload. The Brinell hardness number, HB, is relatedto the applied load and to the surface area of thepermanent impression made by a ball indenter.

e. The Rockwell hardness test is performed inaccordance with the requirements of ASTM E18-92(ASTM 1992a). This test is an indentation hard-ness test to force a diamond spheroconical indenteror hard ball indenter into the surface of the materialin two operations and to measure the difference indepth of the indentation. The Rockwell hardnessnumber, HR, is a number derived from the netincrease in the depth of indentation as the force isincreased from a preliminary test force to a totaltest force and then returned to the preliminary testforce. The higher the number the harder thematerial.

f. The Vickers hardness test is performed inaccordance with the requirements of ASTM E92-82(ASTM 1987a). The Vickers hardness test is anindentation hardness test to force a squarebasedpyramidal diamond indenter with specified faceangles into the surface of the material to measurethe diagonals of the resulting impression afterremoval of the load. Vickers hardness number isrelated to the applied load and the surface area ofthe permanent impression. The hardness values

from different test methods can be correlatedthrough a conversion chart (ASTM E140-88(ASTM 1988)).

g. For each type of hardness test performed, atleast five indentations should be made for eachregion. The result of the hardness test is the aver-age of the indentations.

h. Applicable documents.

(1) ASTM E10-84: Standard Test Method forBrinell Hardness of Metallic Materials (ASTM1984).

(2) ASTM E18-92: Standard Test Methodsfor Rockwell Hardness and Rockwell SuperficialHardness of Metallic Materials (ASTM 1992a).

(3) ASTM E92-82: Standard Test Method forVickers Hardness of Metallic Materials (ASTM1987a).

(4) ASTM E110-82: Standard Test Methodfor Indentation Hardness of Metallic Materials byPortable Hardness Testers (ASTM 1987b).

8. Fracture Toughness Test

Fracture toughness is a material property whichindicates its resistance to fracture. Fracture tough-ness testing provides a measure of resistance tocrack initiation or propagation. Test methodsinclude Charpy V-notch test (CVN), Plane-StrainFracture Toughness test (KIc), and Crack-TipOpening Displacement test (CTOD). The CVN testis used to measure the ability of a material toabsorb energy. TheKIc or CTOD tests are used todetermine critical crack size that a material cantolerate without fracture when loaded to a specificstress level. The welding process and weldingprocedure have a significant effect on the fracturetoughness of a weld joint. The same welding pro-cess and procedure must be used for the structureand test specimens. Fracture toughness test speci-mens should be selected from a distressed gatemember or connection so that the test results arerepresentative of the gate. As an alternative, testsamples may be made of similar materials andwelding procedures to that used in the originalfabrication. Orientations of the test specimenstaken from gate samples should follow the

5-4

ETL 1110-2-34630 Sep 93

provisions specified by AWS B4.0-85 (AWS 1985),Figure D-3. Test specimens should not containmetal that has been affected thermally as a result ofcutting or preparation nor welding stops or starts.The weld metal width-to-specimen thickness rela-tionship provisions are given in AWS B4.0-85,Figure D-4. When an evaluation of the base metalor heat affected zone is required, the location of thenotch should be specified.

a. Charpy V-notch test.

(1) The CVN test provides information aboutbehavior of metal when subjected to a single appli-cation of a load resulting in multiaxial stressesassociated with a notch coupled with high rates ofloading. For some materials and temperatures,impact tests on notched specimens have been foundto predict the likelihood of brittle fracture betterthan tension tests or other tests used in materialspecifications.

(2) The specimen preparation and test proce-dure for the CVN test is described by ASTME23-92 (ASTM 1992b). When specified, the sur-face finish of the V-notch of the Charpy impactspecimen is 20 µin., or less. The testing machine isa pendulum type of rigid construction and of capac-ity more than sufficient to break the specimen inone blow. The test is performed at various speci-fied temperatures.

(3) Five specimens should be tested for eachtest condition and the amount of energy absorbedby the specimen at fracture should be recorded.The highest and lowest values are discarded, andthe result is taken as the average of the remainingthree specimens tested. If any specimen fails tobreak or jams in the machine, the data of that speci-men is not included in the calculation of theaverage.

(4) In addition to the absorbed energy, othertest indicators, such as lateral expansion of thefractured specimen and appearance of the fracturedsurfaces, can also be used to characterize thefracture toughness of the test material. The amountof expansion on each side of each half can bemeasured using a lateral expansion gage. The twobroken halves must be measured individually andthe larger value is used.

(5) The fracture appearance can be quantifiedby measuring the length and width of the cleavageportion of the fracture surface or comparing theappearance of the fractured surface with a fractureappearance chart (ASTM E23-92 (1992b)).

b. Plane-strain fracture toughness test.

(1) The propertyKIc characterizes the resis-tance of a material to fracture in the presence of asharp crack under severe tensile stress. This valuemay be used to estimate the relation between failurestress and defect size for a material in servicewherein the conditions of high tensile stress wouldbe expected. The values ofKIc can be used forinspection and discontinuity assessment criteria,when used in conjunction with fracture mechanicsanalyses.

(2) The plane-strain fracture toughness can beexperimentally determined using compact tensiontest specimen or bend test specimen. The specimenpreparation and test procedures must be in accor-dance with ASTM E399-90 (ASTM 1990g), FiguresA4-1 and A3-1, respectively. For a result to beconsidered valid, it is required that both the speci-men thickness and the crack length exceed2.5(KIc/σys), whereσys is the 0.2-percent offset yieldstrength andKIc is the fracture toughness of thematerial at test temperature and loading rate. Theinitial selection of a size of specimen may be basedon an estimated value ofKIc for the material to betested.

(3) The ASTM requirement for plane-straincondition can be expressed in terms of Irwin’splane-strainβIc value (ASTM E399-90 (ASTM1990g)) as follows:

(5-1)βIc

1t

KIc

σys

2

≤ 0.4

where

t = thickness

σys = material yield strength

If βIc is 0.4, or less, the specimen size is sufficientlylarge to ensure plane-strain behavior and LEFM can

5-5

ETL 1110-2-34630 Sep 93

be applied. Otherwise, elastic-plastic fracturemechanics (EPFM) must be employed in the frac-ture analysis. The crack-tip opening displacement,as discussed in the following section, is usually thematerial toughness parameter for EPFM assessment.

c. Crack-tip opening displacement test.

(1) CTOD is the displacement of the cracksurfaces normal to the original (unloaded) crackplane at the tip of the fatigue precrack. The CTODvalues vary with material toughness depending uponthe amount of plastic deformation at the crack tipunder load. Therefore, CTOD at fracture incipientload indicates the fracture toughness of the testmaterial.

(2) The CTOD values may be used to charac-terize the toughness of materials that are too ductileor lack sufficient size to be tested forKIc. Thedifferent values of CTOD characterize the resistanceof a material to crack initiation and early crackextension at a given temperature. The values ofCTOD can be used for inspection and fractureassessment criteria, when used in conjunction withfracture mechanics analyses.

(3) CTOD tests use three-point bend specimens.Preparation of test specimen and test procedure isdescribed in ASTM E1290-89 (ASTM 1989d). Thecritical CTOD values are derived from measure-ments of load and clip gauge displacement.

d. Applicable documents.

(1) ASTM E23-92: Standard Test Methods forNotched Bar Impact Testing of Metallic Materials(ASTM 1992b).

(2) ASTM E399-90: Standard Test Method forPlane-Strain Fracture Toughness of Metallic Mater-ials (ASTM 1990g).

(3) ASTM E1290-89: Standard Test Methodfor Crack-Tip Opening Displacement (CTOD) Frac-ture Toughness Measurement (ASTM 1989d).

(4) AWS A2.4-86: Standard Symbols forWelding, Brazing, and Nondestructive Examination(AWS 1986).

(5) AWS A3.0-89: Standard Welding Termsand Definitions (AWS 1989).

(6) AWS B4.0-85 - Part D: Fracture Tough-ness Testing of Welds (AWS 1985).

e. CVN-KIc correlations.

Due to ease of testing and cost considerations,CVNtest results are more available thanKIc test results.An approximation ofKIc may be obtained throughthe two-stageCVN-KIc transition method as discus-sed by Barsom and Rolfe (1987).

(1) Determine impactCVN test results in thetransition temperature region at test temperaturesapproximatelyTs above the expected minimum ser-vice temperature,To. Ts is the temperature shift(expressed in degrees Fahrenheit) between fracturetoughness under dynamic loading,KId, and fracturetoughness at slow loading rate,KIc. Transitiontemperatures andTs are described in Enclosure 6.

(5-2)Ts 215 1.5σys

where

Ts = degrees Fahrenheit

σys = yield strength expressed in kips persquare inch

(2) DetermineKId by the followingrelationship

(5-3)KId 5E(CVNIMP )

where

E = Young’s modulus expressed in units ofpounds per square inch

KId = critical stress-intensity factor underdynamic loading (dynamic fracturetoughness) expressed in units ofpounds per square inch-√in.

CVNIMP = ImpactCVN test result in units offoot-pounds

(3) Shift theKId values at each temperature byTs (Equation 5-2) to determine theKIc values as afunction of desired minimum service temperature:KIc(To) = KId(To+Ts).

5-6

ETL 1110-2-34630 Sep 93

Figure 5-1 illustrates the method graphically. Thisprocedure is limited to the lower end of the transi-tion curve, where the impactCVN value in

foot-pounds is less than about one-half of the yieldstrength in kips per square inch.

Figure 5-1. Two-stage CVN-KIc correlation process

5-7

ETL 1110-2-34630 Sep 93

FRACTURE AND FATIGUE EVALUATION

1. Purpose of Evaluation

a. When inspections reveal discontinuities (i.e.cracks or flaws), it is necessary to establish accep-tance levels to determine if repairs are needed toprevent fracture. The critical discontinuity (i.e.defect) size may be determined through a fracturemechanics evaluation for a given set of loads, envi-ronmental factors, geometry, and material proper-ties. If the size of the discontinuity (crack or flaw)is less than the critical defect size, the expectedremaining life and rate of crack propagation may bedetermined by a fatigue analysis. Fracture andfatigue evaluation requires identification of discon-tinuity parameters, material properties, and accep-tance levels. The engineering decision on appropri-ate repair or planned maintenance is based on theconcept of fitness for service of the distressed gatestructure.

b. A lock gate is fit for service when it func-tions satisfactorily during its lifetime without reach-ing any serious limit state. The repair of harmlessdiscontinuities may introduce more harmful, andless easily detectable discontinuities. Repair weldingis often difficult to carry out satisfactorily, since therepair welds are usually made under unfavorableconditions. The needs to repair detected discontinu-ities must be determined in accordance with fitness-for-service concepts.

2. Fracture Behavior of Steel Materials

a. The service temperature under which a lockgate operates has a significant effect on the fracturebehavior of the steel. For low and intermediatestrength steels, the material changes from brittlefracture behavior (i.e.,KIc applies) to ductile frac-ture behavior (i.e.,Kc or CTOD applies) at a certaintransition temperature. This temperature is calledthe nil-ductility transition (also abbreviated as NDTwhich should not be confused with nondestructivetesting) temperature and is measured by the dropweight test (ASTM E208-91 (ASTM 1991i)). TheNDT temperature is defined as the highest tempera-ture at which a standard specimen breaks in a brittlemanner under dynamic loading. At temperaturesabove the NDT temperature, the material has suffi-cient ductility to deflect inelastically before total

fracture. Below the NDT temperature, the fracturetoughness remains relatively constant with changingtemperature. For impact loading, the NDT tempera-ture approximately defines the upper limit of theplane-strain condition as shown in Figure 6-1.

Figure 6-1. Relation between notch toughnessand loading rates (Barsom/Rolfe, FRACTUREAND FATIGUE CONTROL IN STRUCTURES:Applications of Fracture Mechanics,©1987, p 110.Reprinted by permission of Prentice-Hall, Inc.,Englewood Cliffs, NJ.)

b. For steel, the NDT temperature depends onmaterial thickness and applied loading rate. Theanticipated level of structural performance (i.e.brittle or ductile) can be determined from the frac-ture toughness test results performed at tempera-tures around the transition temperature. With anadditional consideration of the geometric constrainteffect due to material thickness (i.e.β factor,Equation 5-1), the appropriate fracture parameterKIc, Kc, or CTOD can be selected for fractureanalysis. For structures subject to static or dynamicloading, the respective fracture toughness-to-temperature relations (i.e.KIc for static loading andKId for dynamic loading) must be used to charac-terize the fracture behavior. Figure 6-1 shows theschematic relationships between level of structuralperformance and service temperature for variousloading rates (Barsom and Rolfe 1987).

3. Fracture Analysis

a. For lock gates operating under the mini-mum service temperature below the nil-ductility

Enclosure 6 6-1

ETL 1110-2-34630 Sep 93

transition temperature, linear elastic fracturemechanics analysis (LEFM andKIc) needs to beperformed for assessing the discontinuities revealedfrom inspections. For gates with discontinuitiesoperating at temperatures above nil-ductility transi-tion temperature, elastic-plastic fracture analysis(EPFM and CTOD) needs to be conducted. In anycase, LEFM may be used as an initial evaluationtool, since it is simple to apply and generally givesa conservative answer. Fatigue analysis is neededwhen the remaining gate life and the crack growthrate are necessary for developing the inspection andmaintenance scheduling for the distressed gates.This section presents a procedure for fracture analy-sis of distressed lock gates.

b. In LEFM analysis, the applied stress-intensity factor (KI) shall always be less than thecritical stress-intensity factor (KIc). The criticaldiscontinuity size is related to material fracturetoughness (critical stress-intensity factorKIc, KId, orKc) for a given applied load and loading rate at theminimum service temperature.

(6-1)acr

1FS

KIc or KId or Kc

Cσ

2

where

acr = critical discontinuity size, inches

KIc (or KId or Kc) = fracture toughness of the gatematerial, kips per squareinch-√in.

σ = applied nominal stress in kips per square inch

FS = factor of safety (e.g. 2)

C = constant which is a function of discontinuityand joint geometry and loading type as shownin Figures 6-8 through 6-16

c. The procedure of fracture assessment ofdiscontinuities may be described by the followingsteps and the flow chart is shown in Figure 6-2:

(1) Determine the actual shape, location, andsize of the discontinuity by NDT inspection.

(2) Determine the effective discontinuitydimensions to be used for analysis (British Standard

Institution 1980, Burdekin et al. 1975, and ASME1978). Discontinuities are classified as throughthickness (may be detected from both surfaces),embedded (not visible from either surface), or sur-face (may be observed on one surface) as illustratedin Figure 6-3. To determine the effective dimen-sions of a discontinuity(ies):

(a) Resolve the discontinuity(ies) into a planenormal to the principal stresses as shown inFigure 6-4. Effective dimensions for various iso-lated discontinuity types are shown in Figure 6-3.

(b) Check interaction with neighboring discon-tinuities to obtain the idealized discontinuity dimen-sions; idealizations for interaction of discontinuitiesare shown in Figures 6-5 and 6-6.

(c) Check interaction with surfaces by recate-gorization as shown in Figure 6-7 for surface orembedded discontinuities (idealized or actual).

(d) Determine final idealized effective dimen-sions for fracture analysis.

(3) Determine the stress level by an appropri-ate structural analysis, assuming no crack exists.Structural loading can be divided into primarystresses,σp, and secondary stress,σs. The primarystress consists of membrane stresses,σm, and bend-ing stressσb, due to imposed loading. Examples ofsecondary stresses include stress increase due tostress concentration imposed by geometry of thedetail under consideration, thermal, and residualstress. For discontinuities at nonheat treated welds,the residual tensile stress should be taken as theyield stress. An estimate of the residual stressshould be used for postheat treated weldments. Theapplied stress is the sum of primaryσp and second-ary σs stresses. If the applied stress is greater thanthe yield stress, EPFM must be employed. Ifapplied stress is less than the yield stress and theplane-strain factorβ < 0.4 (Enclosure 5, paragraph8), LEFM should be used based onKIc. When theapplied stress is less than the yield stress andβ >0.4, Kc (a function of plate thickness) should beused instead ofKIc, if available, otherwise, EPFMbased on CTOD analysis must be employed.

(4) Determine material properties includingσys, E andKIc (based on the level of applied stressand the value ofβIc), Kc, or CTOD. KIc may be

6-2

ETL 1110-2-34630 Sep 93

Figure 6-2. Fracture and fatigue assessment procedure

6-3

ETL 1110-2-34630 Sep 93

estimated from CVN test values by the transition

Figure 6-3. Required dimensions of a discontinuity (after British Standards Institution 1980)

Figure 6-4. Resolution of a discontinuity (afterBritish Standards Institution 1980)

method (Enclosure 5, paragraph 8) if directKIc testdata are not available.

(5) Perform fracture assessment to determinethe critical discontinuity size.

(6) If the discontinuity is noncritical, determinethe remaining life using a fatigue analysis in thisenclosure, paragraph 6.

These steps are further discussed in the followingsections.

d. Fracture mechanics may be used to establishacceptance levels for various discontinuities bycomparing the discontinuity(ies) size with the criti-cal discontinuity (defect) size. Each case is uniquedepending on a given set of loads, environmentalfactors (e.g. temperature), geometry, and materialproperties. The critical discontinuity size is deter-mined using fracture mechanics principles whichrelate stress, discontinuity size, and fracture tough-ness to existing conditions. The stress-intensity

Figure 6-5. Interaction of coplanar discontinu-ities (Extracts from PD 6493: 1980 are repro-duced with the permission of BSI. Completecopies of the standard can be obtained by postfrom BSI Publications, Linford Wood,Milton Keynes, MK14 6LE)

factor KI or CTOD should always be less than thecritical stress-intensity factorKIc , Kc , or the criticalCTOD valueδcrit , respectively.

6-4

ETL 1110-2-34630 Sep 93

Figure 6-6. Interaction of noncoplanardiscontinuities

Figure 6-7. Interaction of discontinuities withsurfaces

4. Linear-Elastic Fracture Mechanics

a. Fundamental concepts of LEFM aredescribed by Barsom and Rolfe (1987). LEFM isvalid only under plane-strain conditions, whenβIc ≤ 0.4. The basic principle of LEFM is thatincipient crack growth will occur when the stress-intensity factorKI (the driving force) equals orexceeds the critical stress-intensity factorKIc (theresistance).KI characterizes the stress field in frontof the crack and is related to the nominal stressσand crack dimensiona for a given load rate andtemperature by

(6-2)KI Cσ a

where

C = dimensionless correction factor for a givengeometry

If C is known,KI can be computed for any combi-nation ofσ anda. Stress-intensity factors for vari-ous types of geometries can be calculated using theinformation included in Figures 6-8 through 6-16(Barsom and Rolfe 1987). Barsom and Rolfe andTada, Paris, and Irwin (1985) contain compilationsof solutions for a wide variety of configurations.After the stress-intensity factor is determined byEquation 6-2, it should be compared to the criticalstress-intensity factor,KIc (determined as describedin Enclosure 5, paragraph 8). An FS = 2.0 appliedto crack length is considered appropriate to preventfracture. Therefore, the crack is considered to beacceptable ifKI < KIc/√2. To determine the allow-able maximum crack size or nominal stress for agiven KIc, substituteKIc for KI and solve fora or σusing Equation 6.2. The critical discontinuity size astructural member can tolerate at a given stressσandKIc with FS of 2.0 is:

(6-3)acr

12

KIc

Cσ

2

b. An approximate method to account forstress gradients is to linearize the stress distribution,and divide it into membrane stressσm and bendingstressσb. The stress-intensity factor for each com-ponent of stress can be calculated separately andthen added together. The total applied stress (σp

andσs) can be linearized and resolved intoσm andσb as shown in Figure 6-17.

5. Elastic-Plastic Fracture Assessment

Rearranging Equation 5-1 (Enclosure 5, paragraph8b(3)), the upper limit of plane-strain behavior maybe determined as

(6-4)KIc

σys

t2.5

6-5

ETL 1110-2-34630 Sep 93

Figure 6-8. Through-thickness crack (copyrightASTM. Reprinted with permission)

Figure 6-9. Double-edge crack (Barsom/Rolfe,FRACTURE AND FATIGUE CONTROL IN STRUC-TURES: Applications of Fracture Mechan-ics,©1987, p 40. Reprinted by permission ofPrentice-Hall, Inc., Englewood Cliffs, NJ.)

6-6

ETL 1110-2-34630 Sep 93

Figure 6-10. Single-edge crack (copyright ASTM.Reprinted with permission)

Figure 6-11. Cracks growing from round holes(copyright ASTM. Reprinted with permission)

6-7

ETL 1110-2-34630 Sep 93

Figure 6-12. Cracks growing from elliptical holes(Barsom/Rolfe, FRACTURE AND FATIGUE CON-TROL IN STRUCTURES: Applications of FractureMechanics,©1987, p 43. Reprinted by permissionof Prentice-Hall, Inc., Englewood Cliffs, NJ.)

When this upper limit is exceeded, extensive plasticdeformation occurs at the crack tip (crack tip blunt-ing) and a nonlinear EPFM model must be used foranalysis. (LEFM analysis usingKc may be used ifthe applied stress is less than yield stress.) Crackgrowth criteria for nonlinear fractures can bemodeled by an R-curve, J-integral, or CTOD analy-sis (Barsom and Rolfe 1987). The CTOD methodis the recommended method of EPFM analysis forevaluating steel lock gates. The recommendedprocedure for cases where the applied stress(σp + σs) is greater than the yield stress (BritishStandards Institution 1980) is as follows:

a. Determine the effective discontinuityparametera. This is the equivalent through thick-ness dimension which would yield the same stressintensity as the actual discontinuities under the sameload.

(1) For through-thickness discontinuities,a = /2.

(2) For surface discontinuities,a is deter-mined by Figure 6-18.

(3) For embedded discontinuities,a is deter-mined by Figure 6-19.

b. Determine allowable discontinuity parame-ter am which is calculated by:

(6-5)am C

δcrit

εy

where

εy = yield strain of the material

δcrit = critical CTOD (determined according toEnclosure 5, paragraph 8)

C = values determined by Figure 6-20

In determination ofC, if the sum of primary andsecondary stresses, excluding residual stress, is lessthan 2σys, the total stress ratio (σp + σs)/σys (includ-ing residual stress) is used as the abscissa inFigure 6-20. If this sum exceeds 2σys, an elastic-plastic stress analysis should be carried out to deter-mine the maximum equivalent plastic strain whichwould occur in the region containing the discontinu-ity if the discontinuity were not present. The valueof C may then be determined using the strain ratio,ε/εy as the abscissa in Figure 6-20.

c. If the effective discontinuity parametera issmaller than the allowable discontinuity parameteram, then the discontinuity is acceptable. Using theprocedure described in paragraph 5b, this enclosure,results in an FS of approximately 2.0 in the deter-mination ofam; Figure 6-20 was developed as adesign curve. Therefore, the calculated criticalcrack size would be equal to 2.0am (British Stan-dards Institution 1980).

6. Fatigue Analysis

a. Fatigue is the process of cumulative dam-age caused by repeated cyclic loading. Fatiguedamage occurs at stress-concentrated regions wherethe localized stress exceeds the yield stress of thematerial. After a certain number of cyclic loads,the accumulated damage causes the initiation andpropagation of a crack.

6-8

ETL 1110-2-34630 Sep 93

Figure 6-13. Edge-notched beam in bending (Barsom/Rolfe, FRACTURE AND FATIGUE CONTROL INSTRUCTURES: Applications of Fracture Mechanics,©1987, p 45. Reprinted by permission of Prentice-Hall,Inc., Englewood Cliffs, NJ.)

b. The total fatigue life is the sum of thefatigue crack-initiation life and the fatigue crack-propagation life to a critical size (Barsom and Rolfe1987).

NT = Ni + Np (6-6)

where

NT = total fatigue life

Ni = initiation life

Np = propagation life

c. All steels have microscopic discontinuities,and welded structures always contain larger discon-tinuities due to the welding process. Thus, the mainconcern in fatigue assessment of welded structuresis to determine the crack-propagation life beforereaching the critical crack size which results inbrittle fracture. The life of a structural componentwhich contains a crack is governed by the rate ofsubcritical crack propagation.

d. Fatigue analysis methods described in para-graphs 7 and 8 are based on extensive analyses oftest results from numerous specimens. Variation intest data is large, and inherent uncertainty exists indefining load and strength parameters. Therefore,

fatigue life predictions should be used as a means toevaluate a reliable service life, not to actually pre-dict when a structure will fail.

7. Fatigue Crack-Propagation (Barsom andRolfe 1987)

The fatigue crack-propagation behavior for metals isshown in Figure 6-21. Figure 6-21 is a plot (log10

scale) of the rate of fatigue crack growth per cycleof load (da/dN) versus the variation of the stress-intensity factor (∆KI). The parametera denotescrack length,N the number of cycles, and∆KI thestress-intensity factor range,KImax to KImin. Based onFigure 6-21, fatigue-crack behavior for steel can becharacterized by three regions:

a. Region I: In region I, for levels of∆KI

below a certain threshold, cracks do not propagateunder cyclic stress fluctuations. Conservative esti-mates of fatigue threshold,∆Kth, can be determinedby

∆Kth = 6.4(1 - 0.85R) ksi-√in. for R > 0.1

∆Kth = 5.5 ksi-√in. for R < 0.1 (6-7)

6-9

ETL 1110-2-34630 Sep 93

Figure 6-14. Embedded elliptical or circular crack (Barsom/Rolfe, FRACTURE AND FATIGUE CONTROL INSTRUCTURES: Applications of Fracture Mechanics,©1987, p 47. Reprinted by permission of Prentice-Hall,Inc., Englewood Cliffs, NJ.)

6-10

ETL 1110-2-34630 Sep 93

Figure 6-15. Surface crack (Barsom/Rolfe, FRACTURE AND FATIGUE CONTROL IN STRUCTURES: Appli-cations of Fracture Mechanics,©1987, p 48. Reprinted by permission of Prentice-Hall, Inc.,Englewood Cliffs, NJ.)

where

R = stress ratio (i.e. fatigue ratio)expressed as

R = σmin / σmax (6-8)

Residual stress should be considered for a cracknear weld area. If∆KI is less than∆Kth, cracks donot propagate.

b. Region II: The fatigue crack-propagationbehavior for∆KI > ∆Kth in region II (i.e. linearportion of the plot on Figure 6-21) may be repre-sented by

da/dN= 3.6x10-10 (∆KI)3 (6-9)

for ferrite-pearlite steels

and

da/dN= 0.66x10-8 (∆KI)2.25 (6-10)

for martensitic steels

(1) For Equations 6-9 and 6-10,a is in unitsof inches, and∆KI in units of kips per squareinch-√in. ASTM A36-91 and A572-91 (1991b andd) Grade 50 steels are classified as ferrite-pearlitesteels, while ASTM A514-91/517 (ASTM 1991cand 1990b, respectively) steels are martensiticsteels. The above equations were based on analysesin air, at room temperature.

6-11

ETL 1110-2-34630 Sep 93

Figure 6-16. Cracks with wedge forces (Barsom/Rolfe, FRACTURE AND FATIGUE CONTROL IN STRUC-TURES: Applications of Fracture Mechanics,©1987, p 52. Reprinted by permission of Prentice-Hall, Inc.,Englewood Cliffs, NJ.)

Figure 6-17. Linearization of stresses (Extracts from PD 6493: 1980 are reproduced with the permission ofBSI. Complete copies of the standard can be obtained by post from BSI Publications, Linford Wood, MiltonKeynes, MK14 6LE)

6-12

ETL 1110-2-34630 Sep 93

Figure 6-18. Relation between dimensions of a discontinuity and the parameter a for surface discontinu-ities (Extracts from PD 6493: 1980 are reproduced with the permission of BSI. Complete copies of thestandard can be obtained by post from BSI Publications, Linford Wood, Milton Keynes, MK14 6LE)

6-13

ETL 1110-2-34630 Sep 93

Figure 6-19. Relation between dimensions of a discontinuity and the parameter a for embedded disconti-nuities (Extracts from PD 6493: 1980 are reproduced with the permission of BSI. Complete copies of thestandard can be obtained by post from BSI Publications, Linford Wood, Milton Keynes, MK14 6LE)

6-14

ETL 1110-2-34630 Sep 93

Figure 6-20. Values of constant C for differentloading conditions (Extracts from PD 6493: 1980are reproduced with the permission of BSI. Com-plete copies of the standard can be obtained bypost from BSI Publications, Linford Wood, MiltonKeynes, MK14 6LE)

Figure 6-21. Fatigue-crack growth in steel

(2) Extensive fatigue-crack growth rate datafor weld metals and heat-affected-zones (HAZ)show that the fatigue rate in weld metals and HAZare equal to or less than that in the base metals.Thus, the above equations can be used for conserva-tive estimates of fatigue-crack growth rates in basemetals, weld metals, and HAZ’s.

c. Region III: Region III is characterized by asignificant increase in the fatigue-crack growth rateper cycle over that predicted for Region II. At acertain value of∆KI, the crack growth rate acceler-ates dramatically. For materials of high fracturetoughness, the stress-intensity factor range valuecorresponding to acceleration in the fatigue-crackgrowth rate (i.e. transition from Region II toRegion III) for zero to tension loading can bedetermined by

KT = 0.04 (E σys)1/2 (6-11)

When theKIc of the material is less thanKT, accel-eration in the fatigue rate occurs at a stress-intensityfactor value slightly belowKIc. Due to the acceler-ation in crack growth rate, a significant increase infracture toughness of a steel aboveKT may have anegligible effect on total fatigue life. Additionally,extrapolation of Region II behavior to Region IIImay overestimate the total fatigue life significantly.

8. Fatigue Assessment Procedures

The procedure to analyze Region II crack growthbehavior in steels and weld metals using fracturemechanics concepts as recommended by Barsomand Rolfe (1987) is as follows:

a. On the basis of the inspection data, deter-mine the maximum initial discontinuity sizeao

present in the member being analyzed and theassociatedKI.

b. Knowing KIc and the nominal maximumdesign stress, calculate the critical discontinuitysize,acr (Equation 6-1), that would cause failure bybrittle fracture.

c. Determine fatigue crack growth rate fortype of steel (Equations 6-9 and 6-10) (i.eferrite-pearlite or martensitic steel).

6-15

ETL 1110-2-34630 Sep 93

d. Determine∆KI using the appropriate expres-sion for KI, the estimated initial discontinuity sizeao, and the range of live load stress∆σ (i.e. cyclicstress range). For cases of variable amplitude load-ing, a spectrum of various discrete stress ranges∆σi

exists. In these cases, an effective stress range∆σe

should be used in determining∆K I. ∆σe can becalculated as the root-mean-cube of the discretestress ranges∆σi,

(6-12)∆σe

m

i 1

ni (∆σi)3

N

1/3

where

ni = number of cycles corresponding to∆σi

N = total number of cycles considered

m = number of discrete stress ranges considered

A live load stress range∆σ, which is due to cycliccompression stresses, may be detrimental in regionswhere tensile residual stress exists. In theseregions, cracks may propagate, since the addition oftensile residual stresses will result in an appliedstress range of tension and compression.

e. Integrate the crack growth rate expression(i.e. Equations 6-9 and 6-10) between the limits ofao (at the initialKI) andacr (at KIc) to obtain thelife of the structure prior to failure. To identifyinspection intervals, integration may be applied withthe upper limit being tolerable discontinuity sizeat.An arbitrary safety factor based on analysis uncer-tainties may be applied toacr to obtainat (forexample, FS = 2.0 was used in Equation 6-3).Another consideration to specifying a tolerablediscontinuity size is crack growth rate. Theat

should be chosen so thatda/dN is relatively smalland a reasonable length of time remains before thecritical size is reached.

f. For a determination ofao.

(1) See Figure 6-3a for through-thicknessdiscontinuities.

(2) For embedded discontinuities (Figure 6-3b),assume that the discontinuity grows until it reachesa circular shape (b= /2). Subsequently, it grows

radially and eventually protrudes a surface at whichtime it should be treated as a surface discontinuityof length .

(3) See Figure 6-3c for surface discontinuities.Initial propagation will result in a semicircularshape. Further propagation will result in the dis-continuity reaching the other surface at which timeit should be treated as a through thicknessdiscontinuity.

9. Development of Inspection Schedules

Inspection schedules can be developed from cracklength versus fatigue life curves. Figure 6-22shows a typical crack length-fatigue life (a - N)curve, which can be obtained from Equation 6-9or 6-10. Critical crack length is determined basedon KIc and maximum design stress as discussed inparagraph 8, this enclosure. The time when repairis needed can be determined considering FS, i.e.,ar = acr/(FS). Remaining loading cycles beforerepair are then determined fromai andar using ana-N curve as shown in Figure 6-22. Inspectionintervals for a gate can be determined from theremaining fatigue life of the members (Common-wealth of Pennsylvania, Department of Transporta-tion 1988).

Figure 6-22. Development of maintenanceschedule

6-16

ETL 1110-2-34630 Sep 93

10. Example Fracture Analysis

Cracks of various shapes were revealed on twotension members on a lock gate by NDT inspection.One member has the cross-sectional dimensions of4 in. thick by 12 in. wide. The other member is1 in. thick by 12 in. wide. The crack types andshapes include: a) single-edge crack; b) through-thickness center crack; c) surface crack along the12-in. side (a/2c = 0.1 and 0.2), and d) embeddedcircular cracks.

The material properties at the minimum servicetemperature of 30oF were determined by materialtestings and are summarized as follows:

σys = 50 ksi σult = 80 ksi

E = 30,000 ksi KIc = 60 ksi-√in.

KId = 40 ksi-√in. δcrit = 0.002 in. (static)

δcrit = 0.001 in. (dynamic)

From structural analysis, the maximum appliedtensile stress is 30 ksi. For each cracked member,the critical crack size will be determined for eachcracking condition under static loading and dynamicloading, respectively:

a. Example for 4- by 12-in. plate:

βIc

1t

KIc

σys

2

14

6050

2

0.36 (Equation 51)

βIc<0.4; therefore, LEFM is applicable.

(1) Single-edge crack (see Figure 6-10).

KI 1.12σ πa k

ab

C 1.12 π k

ab

in Equation 6 2

assumek

ab

1.0

acr

1π

KIc

1.12σ

2

1.02 in.

(Equation 6 1 with no FS)

a/b = 0.17 and = 1.06; therefore,k

ab

iteration is needed foracr and . Afterk

ab

iteration,acr = 0.92 in. ( = 1.05)k

ab

With FS = 2.0,acr = 0.5 (0.92) = 0.46 in.

for dynamic loading:

acr

0.5π

KId

1.12σ

2

0.23 in.

(2) Through-thickness center crack(Figure 6-8).

KI σ πa2bπa

tan

πa2b

assume 2bπa

tan

πa2b

1.0

acr

1π

KIc

σ

2

1.27 in.;

2bπa

tan

πa2b

1.02

After iteration,acr = 1.22 in.With FS = 2.0,acr = 1.22/2 = 0.61 in.

for dynamic loading:

acr

0.5π

KId

σ

2

0.28 in.

(3) Surface crack along the 12-in. side (seeFigure 6-15).

KI 1.12σ π aQ

MK

(a) a/2c = 0.1

6-17

ETL 1110-2-34630 Sep 93

σσys

3050

0.6

from Figure 6-14, Q = 1.02assumeMk = 1.0

acr

Qπ

KIc

1.12σ

2

1.04 in.

(a/t = 0.26;Mk = 1.0)With FS = 2.0,acr = 1.04/2 = 0.52 in.

for dynamic loading:

acr

0.5Qπ

KId

1.12σ

2

0.23 in.

(b) a/2c = 0.2

from Figure 6-14,Q = 1.24assumeMk = 1.0

acr

Qπ

KIc

1.12σ

2

1.26 in.

(a/t = 0.32;Mk = 1.0)With FS = 2.0,acr = 1.26/2 = 0.63 in.

for dynamic loading:

acr

0.5Qπ

KId

1.12σ

2

0.28 in.

(4) Embedded circular crack (see Figure 6-14).

KI σ π aQ

a/2c = 0.5; from Figure 6-14,Q = 2.4

With FS 2.0. acr

0.5Qπ

KIc

σ

2

1.53 in.

for dynamic loading:

acr

0.5Qπ

KId

σ

2

0.68 in.

b. Example for 1- by 12-in. plate:

βIc

1B

KIc

σys

2

11

6050

2

1.44

βIc > 0.4; therefore, EPFM is applicable.

Determine the allowable discontinuity parameteram

(this enclosure, paragraph 5b):

am C

δcrit

εy

(Equation 6 5)

εy

σys

E50

30,0000.0017

σσys

3050

0.6

from Figure 6-20,C = 0.44

For static loading,

am 0.44

0.0020.0017

0.52 in.

For dynamic loading,

am 0.44

0.0010.0017

0.26 in.

Critical crack lengths can be determined for variouscrack shapes from the allowable discontinuityparameteram (this enclosure, paragraph 5a).

11. Example Fatigue Analysis

This example shows how to apply fatigue analysisto determine expected life given an initial flaw size,ai. For this case, consider an initial surface flaw ofthe type shown in Figure 6-15 with a/2c = 0.25.The member is a 4- in.-thick plate of ASTMA572-91 (1991d) Grade 50 steel. The critical stressintensity factor (fracture toughness)KIc of this steelis 60 ksi-√in. at the minimum service temperature.

6-18

ETL 1110-2-34630 Sep 93

The maximum stress level is 30 ksi and the mini-mum stress is 0 ksi. A curve relating the initialsurface flaw sizeai to number of cycles to failureNp will be developed. From Figure 6-15

KI 1.12σ π aQ

MK

σσys

3050

0.6 andQ 1.39 (Figure 614)

assumeMk = 1.0.

With FS 2.0, acr 0.5Qπ

KIC

1.12σ

2

0.71 in.

(for crack sizes up toa = 0.71 in.,Mk = 1.0)for ferrite-pearlite steel,da/dN= 3.6×10-10 (∆KI)

3

(Equation 6-9)

∆KI 1.12∆σ π aQ

50.5 a

Fatigue life can be determined as:

N ⌡⌠acr

ai

da

(3.6×1010)(∆KI)3

N1

(3.6×1010)(50.5)3 ⌡⌠acr

ai

a 3/2 da

N (4.31×104)

1

ai

1

acr

The curve for fatigue life N as a function of initialcrack lengthai for this example is shown inFigure 6-23.

12. Example of Fracture and FatigueEvaluation

a. Single-edge crack.Figure 6-24 shows ahorizontal girder with a single-edge crack. Theinitial crack length is assumed to be 1/8 in.. Theflange plate containing the edge crack is assumed tobe under a cyclic load from zero to maximum ten-sion (i.e., fatigue ratioR = 0). The stress ranges

Figure 6-23. Fatigue life ( N)-initial crack-length(ai) curve

vary from 18 ksi to 27 ksi. The fatigue life can becalculated using the following crack growth equa-tion (Equation 6-9):

dadN

3.6 × 1010(∆KI)3

where

KI 1.12σ πa k

ab

By integrating the crack growth equation, the life ofthe propagating crack can be determined for anycrack length.

N ⌡⌠acr

ai

da

(3.6×1010)(∆KI)3

where

KI = stress-intensity factor which is a functionof crack length

ai = initial crack length

acr =1π

KIc

1.12σk(a/b)

2

(Equation 6 1)

With KIc assumed to be 35 ksi-√in. and a maximumstress of 18 ksi,acr = 0.89 in. using the proceduredescribed in paragraph 10a(1).

6-19

ETL 1110-2-34630 Sep 93

Figure 6-24. A single-edge cracked girder

6-20

ETL 1110-2-34630 Sep 93

Figure 6-25a shows the calculated crack growthversus life cycle for a stress range of 18 ksi(1/2 σys). The remaining lifeN, calculated by theabove equation, is 207,700 cycles. If the structureoperates 10,000 times per year, then the remaininglife of the girder is:

Critical crack length (determined by Equation 6-1)

207,70010,000

20.8 years

is a function of external loading as shown inFigure 6-25b. Figure 6-25c shows the fatigue lifefor stress ranges varying from 18 ksi to 27 ksicalculated using the crack growth equation withvariable stress andacr. The remaining life of thegirder flange containing a 1/8-in. initial crack isshown in the figure as a function of stress.

b. Double-edge crack.A girder flange contain-ing double-edge cracks is shown in Figure 6-26.The crack growth curves were calculated for stressranges varying from 10 to 20 ksi. The same inte-gration procedure as used for the single-edge crackcase is employed for calculating the fatigue life. A1/8-in. initial crack length is also assumed in thiscase. The predicted crack growth curve for stressrange of 18 ksi is shown in Figure 6-27a. Theremaining life of the girder flange plate for variousstress ranges is also shown in Figure 6-27c.

c. Surface crack.Figure 6-28 shows a crackfor which it is assumed initiated in the diagonalbracing member from a surface crack at the cornerof the bracket. It is assumed that the crack propa-gated through the thickness of the bracing memberand then grew toward the edge of the flange plate.A single-edge crack condition similar to the firstexample case was developed. The fracture andfatigue analysis of this example consists of threepropagation steps.

(1) The first step is to analyze the crack prop-agation of a hemispheric surface crack having aninitial radius of 1/16 in. When the surface crackbreaks through the surface on the other side of theplate (i.e., the radius of hemispheric crack becomesthe same as the plate thickness, 3/8 in.), a through-thickness crack condition is reached.

(2) The second step is to analyze crack growthof a plate containing a through-thickness crack.Once the through-thickness crack reaches the edge

Figure 6-25. Curves for fatigue life of a girderwith a single-edge crack

of the plate, the single-edge crack condition isdeveloped.

(3) The third step is to analyze crack growthof the edge crack. The total remaining life of the

6-21

ETL 1110-2-34630 Sep 93

Figure 6-26. A double-edge cracked flange

6-22

ETL 1110-2-34630 Sep 93

Figure 6-27. Curves for fatigue life of a flangewith a double-edge crack

diagonal bracing member from the initial hemi-spheric surface crack can be determined by addingthe three propagation lives. The calculated crackgrowth curve for a stress range of 18 ksi is shownin Figure 6-29a. The total remaining life and criti-cal crack length are also shown in Figures 6-29b,and c, for stress ranges varying from 10 to 20 ksi.

13. Recommended Solutions

a. The recommended solutions to the crackingproblems can be addressed in short- and longtermsolutions. The short-term solution is to repair thefractured members using qualified welding proce-dures and improved fatigue details or bolted coverplates. This temporary measure will ensure contin-uous operation of the structure without catastrophicfailure.

b. A long-term solution will involve detailedinspection and evaluation of the critical membersand connections. Structural analysis using a finiteelement model may be necessary to identify thecritical structural members and connections. Vari-ous loading conditions determined during the courseof previous activities need to be considered in theanalysis. The fatigue category of various weldedconnections should be assessed according to theAmerican Association of State Highway and Trans-portation Officials (AASHTO) Standard Specifica-tion for Highway Bridges (AASHTO 1989) or theANSI/AWS D1.1-92 Structural Welding Code(ANSI/AWS 1992). The expected life of the criti-cal connections can be estimated in accordance withthe respective fatigue category.

c. To maintain satisfactory performance of astructure, a maintenance plan needs to be devel-oped. This maintenance plan should include peri-odic inspections and evaluations. For the worstloading situation, the maximum stress range can bepredicted from an appropriate structural analysis.The inspection intervals can be determined from acrack growth curve of maximum stress range. Theinspection intervals shall be a fraction of theremaining life cycles of the critical members andconnections. These fraction life cycles shall corre-spond to a crack size less than one-half of the criti-cal crack length (i.e. FS = 2.0).

d. Recommended inspection intervals may becomputed using fatigue principles as described inparagraph 9, this enclosure. Using the examplefound in paragraph 12a of this enclosure, theinspection schedule can be determined from thefatigue life curve of the single-edge crack in theprimary member. The maximum stress range isassumed as 18 ksi. The procedure is shown in thefollowing steps.

6-23

ETL 1110-2-34630 Sep 93

Figure 6-28. A stiffening member with a crack

6-24

ETL 1110-2-34630 Sep 93

Figure 6-29. Curves for a fatigue life of a stiffening member with asurface crack

6-25

ETL 1110-2-34630 Sep 93

(1) Determine critical crack length:

acr = 0.89 in. (this enclosure, paragraph 12a)(2) Determine crack length when repair is

needed (Figure 6-22):

ar = 0.89/2 = 0.45 in. (FS = 2.0)

(3) Determine fatigue life from fatigue lifeNversus crack lengtha curve:

N = 160,000 cycles

160,000/10,000 = 16 years (10,000 cycles/year)

Therefore, the girder should be inspected within16 years after the initial crack (ai = 1/8 in.) wasfound.

6-26

ETL 1110-2-34630 Sep 93

REFERENCES FOR ENCLOSURES

a. American Association of State Highway andTransportation Officials. (1989). "Standard specifi-cation for highway bridges," 14th ed., Washington,DC.

b. American National Standards Institute.(1985). "Surface texture," Designation: ANSIB46.1-85, New York.

c. American National Standards Institute/American Welding Society. (1986). "Guide for thenondestructive inspection of welds," Designation:ANSI/AWS B1.10-86, Miami, FL

d. American National Standards Institute/American Welding Society. (1988). "Standard andguide for qualification and certification of weldinginspectors," Designation: ANSI/AWS QC1-88,Miami, FL.

e. American National Standards Institute/American Welding Society. (1992). "Structuralwelding code-steel," Designation: ANSI/AWSD1.1-92, Miami, FL.

f. American Society of Mechanical Engineers.(1978). "Rules for inservice inspection of nuclearpower plant components," ASME Boiler and Pres-sure Vessel Code, Section XI, 1977 ed., New York.

g. American Society for NondestructiveTesting. (1980). "Recommended practice No.SNT-TC-1A," Columbus, OH.

h. American Welding Society. (1985). "Stan-dard methods for mechanical testing of welds,"Designation: B4.0-85, Miami, FL.

i. American Welding Society. (1986)."Standard symbols for welding, brazing and non-destructive examination," Designation: A2.4-86,Miami, FL.

j. American Welding Society. (1989)."Standard welding terms and definitions," Designa-tion: A3.0-89, Miami, FL.

k. American Society for Testing and Materials.(1984). "Standard test method for Brinell hardness

of metallic materials," Designation: E10-84,Philadelphia, PA.

l. American Society for Testing and Materials.(1987a). "Standard test method for Vickers hard-ness of metallic materials," Designation: E92-82(Reapproved 1987), Philadelphia, PA.

m. American Society for Testing andMaterials. (1987b). "Standard test method forindentation hardness of metallic materials byportable hardness testers," Designation: E110-82(Reapproved 1987), Philadelphia, PA.

n. American Society for Testing andMaterials. (1988). "Standard hardness conversiontables for metals," Designation: E140-88,Philadelphia, PA.

o. American Society for Testing andMaterials. (1989a). "Standard practice for loadverification of testing machines," Designation:E4-89, Philadelphia, PA.

p. American Society for Testing andMaterials. (1989b). "Standard test methods forchemical analysis of steel, cast iron, open-hearthiron, and wrought iron," Designation: E30-89,Philadelphia, PA.

q. American Society for Testing andMaterials. (1989c). "Standard practice for hole-type image quality indicators used for radiography,"Designation: E1025-84 (Reapproved 1989),Philadelphia, PA.

r. American Society for Testing and Materials.(1989d). "Standard test method for crack-tipopening displacement (CTOD) fracture toughnessmeasurement," Designation: E1290-89,Philadelphia, PA.

s. American Society for Testing and Materials.(1990a). "Standard specification for straight-beamultrasonic examination of steel plates," Designation:A435/A435M-90, Philadelphia, PA.

t. American Society for Testing and Materials.(1990b). "Standard specification for pressure vessel

R-1

ETL 1110-2-34630 Sep 93

plates, alloy steel, high-strength, quenched, andtempered," Designation: A517/A517M-90,Philadelphia, PA.

u. American Society for Testing and Materials.(1990c). "Standard specification for ultrasonicangle-beam examination of steel plates," Designa-tion: A577/A577M-90, Philadelphia, PA.

v. American Society for Testing and Materials.(1990d). "Standard practice for ultrasonic pulse-echo straight-beam examination by the contactmethod," Designation: E114-90, Philadelphia, PA.

w. American Society for Testing and Materials.(1990e). "Standard practice for ultrasonic contactexamination of weldments," Designation: E164-90,Philadelphia, PA.

x. American Society for Testing and Materials.(1990f). "Standard test methods for chemicalanalysis of carbon steel, low-alloy steel, siliconelectrical steel, ingot iron, and wrought iron,"Designation: E350-90, Philadelphia, PA.

y. American Society for Testing and Materials.(1990g). "Standard test method for plane-strainfracture toughness of metallic materials," Designa-tion: E399-90, Philadelphia, PA.

z. American Society for Testing and Materials.(1990h). "Standard test method for controllingquality of radiographic examination using wirepenetrameters," Designation: E747-90,Philadelphia, PA.

aa. American Society for Testing and Materials.(1990i). "Standard guide for controlling the qualityof industrial radiographic film processing," Designa-tion: E999-90, Philadelphia, PA.

bb. American Society for Testing and Materials.(1991a). "Specification for general requirements forrolled steel plates, shapes, sheet piling, and bars forstructural use," Designation: A6/A6M-916,Philadelphia, PA.

cc. American Society for Testing and Materials.(1991b). "Standard specification for structuralsteel," Designation: A36/A36M-91, Philadelphia,PA.

dd. American Society for Testing and Mater-ials. (1991c). "Standard specification for high-yield-strength, quenched and tempered alloy steelplate, suitable for welding," Designation:A514/A514M-91, Philadelphia, PA.

ee. American Society for Testing and Mater-ials. (1991d). "Standard specification for high-strength low-alloy Columbian-Vanadium steels ofstructural quality," Designation: A572/A572M-91,Philadelphia, PA.

ff. American Society for Testing and Mater-ials. (1991e). "Specification for straight beamultrasonic examination of rolled steel structuralshapes," Designation: A898/A898M-91,Philadelphia, PA.

gg. American Society for Testing and Mater-ials. (1991f). "Standard test methods for tensiontesting of metallic materials," Designation: E8-91,Philadelphia, PA.

hh. American Society for Testing and Mater-ials. (1991g). "Standard guide for radiographictesting," Designation: E94-91, Philadelphia, PA.

ii. American Society for Testing and Mater-ials. (1991h). "Standard test method for liquidpenetrant examination," Designation: E165-91,Philadelphia, PA.

jj. American Society for Testing and Mater-ials. (1991i). "Standard method for conductingdrop weight test to determine nil-ductility transitiontemperature of ferritic steels," Designation: E208-91, Philadelphia, PA.

kk. American Society for Testing and Mater-ials. (1991j). "Standard practice for immersedultrasonic examination by the reflection methodusing pulsed longitudinal waves," Designation:E214-68 (Reapproved 1991), Philadelphia, PA.

ll. American Society for Testing and Mater-ials. (1991k). "Standard reference radiographs forappearance of radiographic images as certainparameters are changed," Designation: E242-91,Philadelphia, PA.

R-2

ETL 1110-2-34630 Sep 93

mm. American Society for Testing and Materials.(1991l). "Standard guide for magnetic particleexamination," Designation: E709-91, Philadelphia,PA.

nn. American Society for Testing and Materials.(1992a). "Standard test methods for Rockwell hard-ness and Rockwell superficial hardness of metallicmaterials," Designation: E18-92, Philadelphia, PA.

oo. American Society for Testing and Materials.(1992b). "Standard test methods for notched barimpact testing of metallic materials," Designation:E23-92, Philadelphia, PA.

pp. American Society for Testing and Materials.(1992c). "Standard method for controlling qualityof radiographic testing," Designation: E142-92,Philadelphia, PA.

qq. American Society for Testing and Materials.(1992d). "Standard test method for guided bendtest for ductility of welds," Designation: E190-92,Philadelphia, PA.

rr. American Society for Testing and Materials.(1992e). "Standard method for radiographic exami-nation of weldments," Designation: E1032-92,Philadelphia, PA.

ss. American Society for Testing and Materials.(1992f). "Standard terminology for nondestructiveexamination," Designation: E1316-92, Philadelphia,PA.

tt. Barsom, J. M., and Rolfe, S. T. (1987)."Fracture and fatigue control in structures," 2nd ed.,Prentice-Hall, Englewood Cliffs, NJ.

uu. British Standards Institution. (1980)."Guidance on some methods for the derivation ofacceptance levels for defects in fusion weldedjoints," BS PD6493, London.

vv. Burdekin, F. M., Harrison, J. D.,Kanazawa, T., Mashida, S., Ekwall, B.,Konkoly, T., and Muncher, L. (1975). "Proposedassessment methods for flaws with respect to failureby brittle fracture,"Welding in the World,Document IIW-471-74, Vol 13, No. ½.

ww. Canadian Standards Association. (1917)."Certification of Welding Inspectors," Designation:CSA W178.2, Rexdale ON, Canada.

xx. Commonwealth of Pennsylvania, Depart-ment of Transportation. (1988). "Guidelines forfatigue and fracture safety inspection of bridges,"Bridge Management Systems Division.

yy. Headquarters, Department of the Army.(1993). "Design of hydraulic steel structures,"Engineer Manual 1110-2-2105, Washington, DC.

zz. Headquarters, Department of the Army.(1984). "Lock gates and operating equipment,"Engineer Manual 1110-2-2703, Washington, DC.

aaa. Headquarters, Department of the Army.(1988). "Periodic inspection and continuing evalua-tion of completed civil works structures," EngineerRegulation 1110-2-100, Washington, DC.

bbb. Tada, H., Paris, P. C., and Irwin, G. R.(1985). "The stress analysis of cracks handbook,"Paris Productions, Inc., Saint Louis, MO.

R-3