Advanced Technologies for Developing Countries
September 21-24, 2005
Slavonski Brod, Croatia /
NEW DESIGN OF REPAIR WELDED JOINTS BASED ON MATERIALS HETEROGENEITY
D. Kozak, L.Tóth, Z.Tonković, Gy. Lenkey, M. Čanađija and Sz.Szabolcs
Keywords: weld joint, plastic yield load , finite element analysis, integrity assessment
1.Introduction
The welding technique is today usually used procedure in producing of many engineering components. The weldment is thought of as made of only one material – all weld or base metal, and the performance difference between weld and base metal could be neglected if the difference is less than 10% 2. It is known that the welding codex demand that mechanical properties of the filler material have to be at least the same or better related to the base material properties.
However, recent welding techniques produce tensile properties of the weld metal which may be quite different from those of the base material 3. It was shown that such heterogeneity in the material composition may be sometime advantageously. Many applications exist in the industry, which require the use of multiple materials in a single component, f.i. inhomogeneous multipass weld joint could be successfully used for repair welding or for weld joints where possible cold hydrogen assisted cracking can appear 6. Also, functionally graded materials (FGM) have been strongly developed in last few years using the benefits of the gradual, controlled variation in material compositionfor dissimilar metal weld applications1.
From the practical point of view, it is necessary to assess the carrying capacity of such performed components, especially their fracture resistance. In that sense, several procedures exist, but recently most used for assessing of crack-like defects in joints with mechanical heterogeneity (strength mismatch) is SINTAP procedure 8.If the materials in the joint posses large plasticity behaviour, one of the measure of estimation to fracture is the value of plastic yield load needed for the yielding of material through the whole net section. This value should be applied for the component integrity assessment by using of Failure Assessment Diagram (FAD).
Yield load solutions for the different strength mismatch welded components are given in the Ref. 7, but they offer the expression for the yield load only in the case when the weld joint is homogeneous. The goal of this paper is to present the yield load solutions obtained by plane strain finite element analysis for most used fracture toughness specimen – single edge notch bend SE(B) specimen, which consists two different materials in the weld. The benefits of the approach where the one half of the weld is composed from the strength overmatch (OM) metal and the second half from the strength undermatch (UM) metal will be considered. Practical application of the weld component, where the crack tip can be located in the OM or UM weld part will be discussed. Obtained finite element mismatch yield load solutions in the form of diagrams will be compared with existing solutions in literature for homogeneous either overmatch or undermatch weld.
2.Yield load analysis generally
Existing defect assessment methods for homogeneous structures need to be modified to incorporate the strength mismatch effects [7]. The accuracy of such methods is directly related to accurate estimates of the yield load FY. The failure assessment diagram (FAD) approach uses normalised stress intensity factor and load axes, and the area under the failure assessment diagram represents conditions of safe operation whereas failure is to be expected when the service point lies outside the curve (Fig. 1). Its general formulation Kr = f (Lr), where Kr = K/Kmat represents the applied stress intensity factor normalised with fracture toughness and Lr = F/FY is the applied load normalised with the yield load.
Figure 1. FAD: Basic analysis against fracture toughness
Yield load analysis may be used to determine the failure stress in the presence of a flaw for materials that are highly ductile. This analysis method assumes that the entire cross section of the component becomes fully plastic before the onset of failure (Fig. 2).
Figure 2. Characteristic yielding zone spreading by SE(B) fracture toughness specimen
3.Yield load solution for all-base specimen
Compendium of the limit load solutions offers the equation for the plane strain yield load solution when the whole SE(B) specimen is made from base metal7:
(1)
where ,
YB denotes the yield strength of base metal, B and W are the specimen thickness and width, a stands for crack length and S=4W. In this paper, standard SE(B) specimen has been considered with W=2B cross section, made from high strength steel with yield strength of 545 MPa.
Key geometrical parameters: crack length a and weld root width 2H were systematically varied enabling estimation of the yield load values for the different component configurations. All combinations of configurations have been presented on the Figure 3, where each key-point determines the values of a and H for particular finite element model.
Figure 3. Key-points, which correspond to particular finite element model of SE(B) specimen
4.Yield load solution for the crack located in overmatched halve of heterogeneous weld
The load, which induces the yielding continuously through the ligament, has been observed directly from plane strain finite element analysis. All materials in the joint have been considered as bilinear elastic, with the same exponent of strain hardening after yielding. Standard plane strain 8-node isoparametric finite elements have been used. Having in mind the symmetry of the specimen, only one half was modelled. Singular elements were considered around the crack tip, with the size of 0,1 mm. The detail of typical finite element mesh is presented in the Figure 4.
Figure 4. The detail of FE mesh of the half of SE(B) fracture specimen with H=W/8 and a/W=0,3
The increment of the increasing load had to be very fine in order to precisely estimate the exact moment of material yielding. Mismatch yield load solutions FYM related to the solution for the specimen made from pure base metal FYB for the SE(B) specimen with different weld width 2H, are given in the Figure 5, depending on crack length ratio a/W [5]. It is evident that yield load values are most distinguished for the specimen with deepest crack. They vary from the overmatch solution for specimens with narrow welds to undermatch solution in the case of welds with low slenderness. All configurations with short cracks possess almost the same behaviour, regardless of the weld width.
Figure 5. Comparison of the mis-match yield load solutions: plane strain SE(B) specimen with crack located in OM
5.Yield load solution for the crack located in undermatched halve of heterogeneous weld
All yield load solutions obtained by assuming the crack tip location in the undermatched weld structure have an overmatch character (Figure 6). It is obviously that overmatch region ahead the crack tip plays the dominant role on the complete fracture behaviour [4]. This is worthy to note from the practical point of view. Namely, the welding codex procedure demands the selection of filler material with better mechanical properties than the base material. This procedure could be fairly questionable by repair welding application. Comparing the yield load results for most extreme investigated situation, where the crack has already advanced to the half of specimen width, one can remark that the difference between the configurations with the crack lying in OM and UM is more than 1/3. The difference value depends on yield strength of particular materials in the weldment also.
Figure 6. Comparison of the mis-match yield load solutions: plane strain SE(B) specimen with crack located in UM
6.Results and discussion
This paper compares the yield load values for the SE(B) fracture toughness specimen with the crack tip located either in the over- or undermatched weld part, obtained by using plane strain finite element analysis. All solutions are given in the form of diagrams, where the mismatch yield load was referred to the yield load of all-base metal specimen, varying the most influenced geometrical parameters: crack length ratio a/W and slenderness of the weld (W-a)/H. The yield strengths of particular materials in the joint were kept constant.
It was found that strength mismatch within the weld has very small influence when the specimen with the short crack positioned in the narrow weld was analysed. These yield load solutions are very close to the solutions, which are valid for all-base specimen. On the other hand, highly constrained specimens with a/W=0,5 show very different behaviour, depending on the weld width and material where the crack tip was positioned. The highest yielding resistance was observed by configuration cracked in the undermatched weld part with highest value of slenderness, whilst lowest yielding resistance is present in configuration with crack located in the overmatched material. This investigation shown that material within the crack tip is located does not play dominant role on the yield load solution, but material in front of it. Having in mind unexpected high values of yield loads obtained for the specimen with crack lying in UM region with OM on opposite side, such configuration should be taken as more considerable in design of welded components.
References
[1]Farren, J., Noecker, F.F.II, DuPont, J.N., Marder, A.R., "Direct Fabrication of a Carbon Steel - to - Stainless Steel Functionally Graded Material for Dissimilar Metal Weld Applications",submitted for publication to the Welding Journal
[2]JianqingF. and YaowuS.,"Mechanical heterogeneity and validity of J-dominance in welded joints", International Journal of Fracture 106 (2000), pp. 311-320.
[3]KoçakM. editor,"Proceedings European Symposium on Assessment of Power Beam Welds", GKSS Research Centre Publications Geesethacht, Germany, 1998.
[4]Konjatic P., Kozak D., Gubeljak N., Predan J. and Matejiček F.,"Yield load solution for the SE(B) fracture toughness specimen with heterogeneous weld joint", 3rd M.I.T. Conference on Computational Fluid and Solid Mechanics, Cambridge, USA, 2005 (
[5]Kozak D., Vojvodič-Tuma J., GubeljakN. and Semenski D. ,"Factors influencing the yielding constraint for cracked welded components", Materials and Technology 39 (2005) 1-2, pp. 29-36.
[6]Rak I, Koçak M, Petrovski B,"Fracture evaluation of repair welding joints for offshore application", Glasgow: 12th International Conference OMAE, 1993.
[7]Schwalbe K-H, Kim Y-J, Hao S, Cornec A, Koçak M. EFAM ETM-MM 96,"The ETM method for assessing the significance of crack-like defects in joints with mechanical heterogeneity (strength mismatch)", Geesthacht: GKSS Research Center, GKSS/97/E9, 1997.
[8]SINTAP: "Structural Integrity Assessment Procedure", FinalRevision, EU-Project BE 95-1462, Brite Euram Programme, 1999.
Doc.dr.sc.DražanKozak
University of Osijek, Mechanical Engineering Faculty, Trg Ivane Brlić-Mažuranić 2, HR-35000 Slavonski Brod, Croatia, Tel.+38535446 188, Fax: +38535446 446, e-mail:
Prof. LászlóTóth
Bay Zoltán Foundation for Applied Research, Director of Institute for Logistics and Production Systems, 2. Iglói st., H-3519 Miskolctapolca, Hungary, Phone: +36 46 363 622, Fax: +36 46 422 786, E-mail:
Doc.dr.sc.ZdenkoTonković
University of Zagreb, Faculty of Mechanical Engineering and Naval Architecture, Ivana Lučića 5, HR-10000Zagreb, Croatia, Tel.+385 1 6168 450, Fax: +385 1 6168187, e-mail:
Dr. GyöngyvérB.Lenkey
Bay Zoltán Foundation for Applied Research, Institute for Logistics and Production Systems, Head of the Department for Structural Integrity, 2. Iglói st., H-3519 Miskolctapolca, Hungary, Phone: +36 46 560118, Fax: +36 46 369438, E-mail:
Doc.dr.sc.MarkoČanađija
University of Rijeka, Faculty of Engineering, Vukovarska 58, HR-51000Rijeka, Croatia, Tel.+38551 651496, Fax: +38551 651 490, e-mail:
Mr. SzávaiSzabolcs
Bay Zoltán Foundation for Applied Research, Institute for Logistics and Production Systems, 2. Iglói st., H-3519 Miskolctapolca, Hungary, Phone: +36 46560121, Fax: +36 46 422 786, E-mail: