Welding Research Sponsored by the Welding Research Council SUPPLEMENT TO THE WELDING JOURNAL, DECEMBER 1972 Procedures for Evaluation of Fracture Toughness of Heat-Affected Zones BY L. J. McGEADY A good quality weld requires adequate fracture tough­ ness of base metal, weld metal, and heat-affected zone (HAZ). Reliable methods are available for the first two but the HAZ evaluation has posed particular experi­ mental problems which are dealt with in this paper. Introduction ly low transition temperature. The ceptibility of all weldment com­ World War II ship failure problem is ponents to fracture. Two approaches Welded structures must be viewed an excellent case for reference. are possible. Testing weldments in essentially as three-component com­ In weldments of quenched and tem­ composite specimen procedures such posites of weld metal, plate and heat- pered steels, the potential for ex­ as the Explosion Bulge or Delta affected zone (HAZ). Inadequate tended fractures not only in plate but allows identification of the most frac­ toughness in any one is a threat to also in either HAZ or weld metal must ture prone elements. Procedures to structural integrity. be considered. The steels in this test the toughness of the weld, plate In carbon steel weldments the frac­ group, because of differences in alloy and HAZ represent a second ap­ ture situation is generally a plate content and kind differ widely in plate proach. problem in the main. The protective toughness characteristics and in The Drop-Weight (DW) and Dynam­ geometric aspects of U- or Vee- sensitivity to HAZ damage in welding. ic Tear (DT) tests have been used groove geometries and the toughness At the same time the higher strength widely to characterize plate and weld levels in welds deposited with requirements for weld metals for metal toughness separately and modest control generally eliminate these steels pose considerations in quantitatively. From tests using these the HAZ and weld metal as common respect to toughness which become two specimens, data are presented in paths of major failure. However flaws more acute as advanced strengths or this report to demonstrate the ap­ in either area are capable of exten­ more economic compositions or both plicability of these two procedures to sion to the plate. Therefore the key to are sought. the HAZ as well. Results of composite catastrophic fracture control has cen­ The control of fracture via the plate specimen tests and single component tered on the use of steel of sufficient- transition temperature while possible tests will be shown to be in agree­ for carbon steel weldments is insuf­ ment when applied to several steels. ficient for quenched and tempered Consequently the understanding of L. J. McGeady is professor and head of the steel weldments. Evaluative methods results of composite specimens tests Department of Metallurgical Engineering for the quenched and tempered (Q+T) is enhanced and their applicability at Lafayette College, Easton, Pa. steel must take into account the sus­ better defined. At the same time the WELDING RESEARCH SUPPLEMENT! 569-s successful characterization of HAZ this test, a weld deposited in a double- strengths or more economic composi­ toughness by the DW and DT U-joint was restrained from shrinking tions or both are sought. Optimum procedures provides quantitative ref­ in cooling. Variable restraint allowed properties are realized only by careful erence data for use in the Fracture definition of a threshold value above selection of filler material and appro­ Analysis and Ratio Analysis Dia­ which cooling cracks would form. priate control of the welding process. grams. Either weld metal or HAZ could fail Like the weld metal, plate material depending on the materials and weld­ can be handled as a distinctly sep­ ing conditions. arate weldment component. It is The Weldability Problem in Consideration of the HAZ as an clearly recognized that differences in Relation to Fracture Toughness area of potential difficulty led to vari­ amount and kind of alloyirg elements in Carbon Steels ous researches. Doan, Stout, Tor and lead to differences in toughness char­ 2 The ship failures in World War II Frye proposed a systematic ap­ acteristics at any strength level. typified the weldability problem in proach to controlling hardness and Some of the steels in the Q+T group carbon steel structures. The intensive brittleness in the coarse-grain heat- show neither a sharp transition tem­ investigation of these failures re­ affected zone through understanding perature nor a high energy shelf vealed that the fracture problem was the relations between weld heat while others do. It is evident that in essentially a plate metal problem, input, plate thickness and temper­ Q+T steel weldments the plate re­ that is, high plate transition temper­ ature, and corresponding cooling mains a critical element for fracture atures, since they were the path for rates in the weld area and Jominy control. extended failures. Adequate tough­ test data. Other investigations While the plate metal and HAZ can ness in the weld metal was possible showed potential embrittlement in re­ be viewed as separate entities, the through use of commercial filler ma­ gions of the HAZ susceptible to strain- plate must be regarded as the parent terials deposited under modest con­ aging type phenomena. Still others of the HAZ. Both are subject to the trol. Sufficient compensation for met­ showed embrittlement possible in the same metallurgical considerations. allurgical damage via the protective metallurgical areas heated just above The fracture toughness realizable in geometric aspects of joints with U- the lower critical temperature. each is dependent upon the nature or Vee-groove configurations was ef­ However, the situation with the car­ and level of alloy content. The plate is fective in minimizing extended frac­ bon steel plates remained the same, heat treated for optimum toughness tures in the HAZ. namely the potential for long disas­ in prescribed thermal cycling pro­ cesses under control in the mill. The However flaws in either the weld trous failures in plate with the weld more complex cycling experienced in metal or heat-affected zone could and zone limited to the role of site or the heat-affected zone is subject to did act as sources for fractures which source of initiation. Commercial weld­ less control in welding. It is unrealis­ were transmitted to and extended in ing material and control of welding tic to expect the heat-affected zone to the plate. As a consequence a two- conditions could be relied on to elim­ exhibit the same microstructures pronged attack was brought to bear inate the weld zones as paths for found in the plate. Control of the on the carbon steel fracture problem major fracture. Ultimately it became welding process must be the route to — one, to produce plate with lower clear that the overall approach to frac­ produce HAZ microstructures with op­ transition temperatures, the other to ture safe design with these steels re­ timum properties in a given steel. In identify and eliminate creation of quired the use of sufficiently low tran­ general the hardenability of the steel flaws in welding. sition temperature plate material to prevent extensive fracture. dictates the latitude of welding condi­ In the effort, a host of tests was tions allowable for retention of the developed. In the context of the prob­ most favorable HAZ microstructure lem, all were called weldability tests. Weldability and Fracture with resultant minimal damage. Some were designed to measure Toughness of High-Strength plate toughness more satisfactorily or completely than possible by the Steels Charpy test; these were regarded as Weldability is a more complex prac­ Weldability Tests service tests. Others generally tical matter for the high-strength Among the various procedures pro­ termed fabrication tests were de­ steels than for the carbon steels. Not posed for evaluating fracture perfor­ signed to measure susceptibility of only the plate but also the heat- mance of weldments three funda­ the base plate to heat-affected zone affected zone and the weld metal re­ mentally different approaches can be cracking. One was the bead on plate quire consideration. Each can be recognized. One is to measure the cracking test in which a weld bead viewed as a separate concern. tendency of plate to accept weld ini­ was deposited on the plate material Conceptually and for practical test­ tiated fractures restricted to the trans- and then examined for evidence of ing purposes the weld metal is isola- verse-to-weld path of travel. A sec­ cracking in the HAZ under the natural ble for separate handling. However ond is to characterize the toughness shrinkage stresses generated by weld the fracture toughness problems of weld, plate and HAZ separately. A cooling. there must be considered in relation third is to use composite test proce­ A second procedure was the "clip to the steels being welded. For car­ dures which allow simultaneous load­ test" in which a piece of plate or bon steels the problem is minimal. ing of all three components and a self- "clip" was welded with a single fillet Adequate toughness or toughness selected fracture path in the compo­ weld to form a Tee-joint with the equivalent to that of the plate is read­ nent of least resistance. material being examined. Testing ily obtained in the weld metal with In specimens such as the Kinzel or consisted of striking the clip, which relatively modest attention to the Lehigh Longitudinal Bend tests3 was the vertical bar of the Tee-joint, welding process using readily avail­ which force fracture transverse to the with a sledge to determine whether able filler materials. Welding of high- weld, the specimen is prepared by de­ in breaking it away, fracture took strength steels presents more strin­ posit of a weld bead on a plate sur­ place in the brittle coarse-grained gent demands.