AWS WHC1 06-2001 Test Methods for Evaluating Welded Joints.pdf

1、TEST METHODS FOR EVALUATING WELDED JOINTSPrepared by theWelding Handbook Chapter Committee on Test Methods for Evaluating Welded Joints:D. E. Williams, ChairConsulting EngineerD. M. BeneteauCenterline (Windsor)LimitedJ. A. ClarkWestinghouse ElectricCorporationB. H. LyonsConsultantE. R. SampsonConsul

2、tantR. F. WaiteConsultantWelding Handbook Volume 1 Committee Member:D. W. DickinsonThe Ohio State UniversityContentsIntroduction 2Testing for Strength 3Hardness Tests 18Bend Tests 22Fracture ToughnessTesting 23Fatigue Testing 34Corrosion Testing 39Creep and RuptureTesting 42Testing of ThermalSpray A

3、pplications 43Weldability Testing 46Conclusion 54Bibliography 54SupplementaryReading List 56CHAPTER 92 TEST METHODS FOR EVALUATING WELDED JOINTSINTRODUCTIONAll types of welded structuresfrom steel bridges tojet componentsserve a function. Likewise, the weldedjoints in these structures and components

4、 are designedfor service-related capabilities and properties. Predict-ing service performance on the basis of laboratory test-ing presents a complex problem because weld size,configuration, and the environment as well as the typesof loading to which weldments are subjected differ fromstructure to st

5、ructure. This complexity is furtherincreased because welded jointsconsisting of unaf-fected base metal, weld metal, and a heat-affected zone(HAZ)are metallurgically and chemically heteroge-neous. In turn, each of these regions is composed ofmany different metallurgical structures as well as chem-ica

6、l heterogeneities.Testing is usually performed to ensure that weldedjoints can fulfill their intended function. The ideal test,of course, involves observing the structure in actual orsimulated service. An example of such “mock-up” test-ing is that done to validate new designs of momentframe and simi

7、lar connections for large buildings instrong seismic areas.1Unfortunately, mock-up andactual service tests are expensive, time consuming, andpotentially hazardous. Therefore, standardized testsand testing procedures are performed in the laboratoryto compare a specimens results with those of metalsan

8、d structures that have performed satisfactorily in ser-vice. Standardized testing provides a bridge between the1. American Institute for Steel Construction (AISC), Seismic Provi-sions for Structural Steel Buildings, Chicago: American Institute forSteel Construction.properties assumed by designers an

9、d analysts and thoseexhibited by the actual structure.Mechanical testing provides information on themechanical or physical properties of a small sample ofwelds or metals to infer the properties of the remainingmaterial within a lot, heat range, or welding procedure.Standardized procedures are used t

10、o sample, orient,prepare, test, and evaluate the specimens in order toprovide results that can be compared to design criteria.For example, virtually all design codes are based on aminimum tensile strength that must be achieved notonly in the base metal but also in the weldment.When selecting a test

11、method, the tests purpose mustbe considered and balanced against the amount of timeand the resources available. For example, tension andhardness tests both provide a measure of strength, butthe latter are simpler and more economical to perform.Hardness tests can be used to confirm that adequatestren

12、gth has been achieved in some heat-treated compo-nents. Although they can verify that a maximum heat-affected-zone hardness has not been exceeded, hardnesstests cannot adequately establish the strength of awelded joint because of the heterogeneous nature ofwelds. Regardless of the differences betwee

13、n test meth-ods, all testing procedures measure either a compositeaverage or a “weak-link” component of the property ofinterest within the area sampled. Thus, an understand-ing of the test details is necessary to interpret theresults.When testing a welded or brazed joint, the investiga-tor must not

14、only relate the test to the intended serviceof the actual structure but also determine whether trueproperties are measured by the limited region tested.TEST METHODS FOR EVALUATING WELDED JOINTSCHAPTER 9TEST METHODS FOR EVALUATING WELDED JOINTS 3Test results must therefore be carefully interpreted an

15、dapplied. As each laboratory test provides only a limitedamount of information on the properties of weldedjoints, most weldments are evaluated using several tests.Each test provides specific data on the serviceability ofthe weldment. The properties evaluated by testinginclude strength (e.g., ultimat

16、e tensile strength, yieldstrength, shear strength), tensile ductility (e.g., elonga-tion and reduction of area), bend test ductility, tough-ness (e.g., fracture toughness, crack arrest toughness,and Charpy V-notch toughness), fatigue, corrosion, andcreep. The scope of the testing is either defined a

17、s partof the investigation or specified in the relevant code orstandard, depending on the application.Testing should be performed on samples that reflectthe heat treatment condition used in service. However,the topic of the aging of steel specimens often arises intesting welded joints. In this conte

18、xt, aging is a degas-sing treatment at room temperature or a slightly ele-vated temperature. For example, the American WeldingSocietys filler metal specification for carbon steel fluxcored arc welding electrodes,2as well as some weldingcodes such as Structural Welding CodeSteel, AWSD1.1:2000,3, 4per

19、mit the aging of tension test speci-mens at 200F to 220F (93C to 104C) before testing.However, other codes such as the Bridge Welding Code,ANSI/AASHTO/AWS D1.5-96,5do not permit agingfor weld procedure qualification tests.The welding process can introduce hydrogen into theweld metal, mostly from wat

20、er that is disassociatedunder the high temperature of the arc. The hydrogendiffuses out over time but may introduce anomalies intotensile test results. These can sometimes be seen as“fisheyes” (small pores surrounded by a round, brightarea on the fracture surface of tension tests of steelwelds) even

21、 though normal cup-and-cone fracture maybe observed, if tested only days later, and the yieldstrength, ultimate strength, and impact test results willremain unchanged. Such low-temperature aging is per-mitted because it does not change the metallurgical2. American Welding Society (AWS) Committee on

22、Filler Metals,Specification for Carbon Steel Electrodes for Flux Cored Arc Weld-ing, ANSI/AWS A5.20, Miami: American Welding Society.3. American Welding Society (AWS) Committee on Structural Weld-ing, 2000, Structural Welding CodeSteel, AWS D1.1:2000, Miami:American Welding Society.4. At the time of

23、 the preparation of this chapter, the referenced codesand other standards were valid. If a code or other standard is citedwithout a date of publication, it is understood that the latest editionof the document referred to applies. If a code or other standard iscited with the date of publication, the

24、citation refers to that editiononly, and it is understood that any future revisions or amendments tothe code or standard are not included; however, as codes and stan-dards undergo frequent revision, the reader is encouraged to consultthe most recent edition.5. American Welding Society (AWS) Committe

25、e on Structural Weld-ing, 1996, Bridge Welding Code, ANSI/AASHTO/AWS D1.5-96,Miami: American Welding Society.structure; it simply quickens the diffusion of hydrogenfrom the weldment. With this one exception, weldmenttesting is typically performed using specimens that rep-resent the heat treatment co

26、ndition of the weldment asit will be used in service.The various testing methods used to evaluate theexpected performance of welded and brazed joints andthermal spray applications are examined in this chapter.The description of each method includes a discussion ofthe property being tested, the test

27、methods used, theapplication of results, and, most importantly, the man-ner in which these results relate to welded joints. Anoverview of weldability testing is also presented.6This chapter makes frequent reference to the Stan-dard Methods for Mechanical Testing of Welds, ANSI/AWS B4.0 and AWS B4.0M

28、,7and Standard Methodsand Definitions for Mechanical Testing of Steel Prod-ucts, ASTM A 370.8The latest edition of these stan-dards should be consulted for more information onthe testing and evaluation of welded joints. In addi-tion, the American National Standard Safety inWelding, Cutting, and Alli

29、ed Processes, ANSI Z49.1,9should be consulted for rules regarding health and safetyprecautions.TESTING FOR STRENGTHThe design of nearly every component and structureis based on minimum tensile properties. As weldedjoints contain metallurgical and often compositionaldifferences that result from the w

30、elding process, theeffects of these changes on the mechanical properties ofthe weldment must be assessed. Some strength tests,such as tension tests, measure tensile strength directly,while others, such as the peel test, verify that the weld isas strong as the base metal. The various techniques usedt

31、o evaluate the strength of weldments are discussedbelow.6. Weld soundness is evaluated using the nondestructive examinationmethods described in Chapter 14, Vol. 1 of the Welding Handbook,9th ed., Miami: American Welding Society.7. American Welding Society (AWS) Committee on Mechanical Test-ing of We

32、lds, Standard Methods for Mechanical Testing of Welds,ANSI/AWS B4.0, Miami: American Welding Society; American Weld-ing Society (AWS) Committee on Mechanical Testing of Welds, Stan-dard Methods for Mechanical Testing of Welds, AWS B4.0M, Miami:American Welding Society.8. American Society for Testing

33、 and Materials (ASTM) Subcommit-tee A01.13, Standard Test Methods and Definitions for MechanicalTesting of Steel Products, ASTM A 370, West Conshohocken, Penn-sylvania: American Society for Testing and Materials.9. American National Standards Institute (ANSI) Accredited StandardsCommittee Z49, Safet

34、y in Welding, Cutting, and Allied Processes,ANSI Z49.1, Miami: American Welding Society.4 TEST METHODS FOR EVALUATING WELDED JOINTSTENSION TESTSTension tests are conducted to evaluate the strengthand ductility of the base metal, the weld metal, and theweldment. These tests provide quantitative data

35、for theanalysis and design of welded structures. Tension testsare frequently conducted to verify that the tensilestrength of a weldment meets specified minimum values.FundamentalsTension testing involves the loading of specimens intension until failure occurs. These tests provide infor-mation about

36、two strength valuesyield strength andultimate tensile strengthalong with two measures ofductilityelongation and the reduction of area. Allthese properties are often reported for the base materialfrom tests performed by the manufacturer of the mate-rial. The modulus of elasticity (Youngs modulus) can

37、also be determined from the tensile test, although thisproperty does not vary significantly for a given materialand is not typically reported.Tension Test Specimens. The various types of ten-sion tests differ primarily with respect to the orientationof the test specimen within the weldment or struct

38、ure.For example, in the uniaxial tension test used in weld-ment and base metal evaluation, specimens have eithera circular (“round”) or rectangular (“strap”) cross sec-tion within a reduced gauge length. Two typical speci-men configurations are illustrated in Figure 1.Test Procedure. The specimen en

39、ds, which may beplain, threaded, or grooved depending on the particulargrips used, are placed within the two grips of a tensionmachine. The grips are then moved away from eachother, placing the specimen in uniaxial tension while theload placed on the specimen is monitored or recorded.Only the maximu

40、m load is required to define the ulti-mate tensile strength (UTS) (see below).Data Collection and Interpretation. Simultaneousrecording of both the applied load and the resultingspecimen displacement (i.e., lengthening) within thegauge length is necessary to determine the yield strengthand define th

41、e stress-strain curve. The engineeringstress-strain curve is developed by plotting the instanta-neous load divided by the original cross-sectional areaof the test specimen against the instantaneous displace-ment (within the gauge length) divided by the originalgauge length of the specimen. A stress-

42、strain curve forlow-carbon steel is presented in Figure 2.As can be observed in Figure 2, stress-strain diagramsprovide information on (1) the ultimate tensile strength,(2) the yield point, (3) offset yield strength, and (4) themodulus of elasticity. The percent elongation and per-cent reduction can

43、 also be determined with the test.For engineering applications, the term ultimate ten-sile strength (UTS) is defined as the maximum loaddivided by the original cross-sectional area of the speci-men. The maximum load can be taken directly from thetension testing machine load pointer. Alternatively, i

44、tcan be derived from the peak of the stress-strain curveafter yielding. It should be noted that the maximumload is not typically the load at fracture, which is usu-ally lower because extensive local yielding has signifi-cantly reduced the cross-sectional area of the specimen.The elongation of the sp

45、ecimen is distributed nearlyuniformly over the reduced section until the maximumload is reached, but it becomes increasingly local asnecking occurs.Key:T = Thickness of the reduced section (rectangular), in.(mm)W = Width of the reduced section (rectangular), in. (mm)D = Diameter of the reduced secti

46、on (round), in. (mm)L = Specimen length, in. (mm)R = Radius, in. (mm)P = Parallel (“reduced”) sectionG = Gauge length, in. (mm)Source: Adapted from American Welding Society (AWS) Committee on MechanicalTesting of Welds, 1998, Standard Methods for Mechanical Testing of Welds, ANSI/AWSB4.0-98, Miami:

47、American Welding Society, Figures A11 and A12.Figure 1Typical Tensile Specimens:(A) Rectangular and (B) RoundTEST METHODS FOR EVALUATING WELDED JOINTS 5The yield strength, an arbitrarily defined measure, isintended to indicate the stress beyond which permanentdeformation occurs. The stress-strain cu

48、rve of somematerials exhibits a well-defined upper yield point thatclearly marks the end of fully elastic behavior, andstresses above this point permanently elongate thematerial. The initial straight portion of the curve has aslope equal to the modulus of elasticity. The upper yieldpoint is followed

49、 by yielding at a constant, slightlylower load (the lower yield point), after which the curverises again due to work hardening. However, arbitrarydefinitions of yield strength are necessary because somematerials do not exhibit a clear yield point.The offset yield strength is determined by construct-ing a line that is parallel to the elastic, straight part ofthe curve (i.e., with a slope of the modulus of elasticity)but offset along the strain axis by 0.2% strain. Thestress at the intersection of this offset line and the stress-strain curve is the 0.2% offset yield strength. T