Researchonfatiguebehaviorofweldedjointsprayingfusedbylowtransformationtemperaturealloypowder Xiaohui Zhao a, ⇑ , Dongpo Wang b , Caiyan Deng b a Key Laboratory of Automobile Materials, School of Materials Science and Engineering, Jilin University, Changchun 130025, China b School of Materials Science and Engineering, Tianjin University, Tianjin 300072, China article info Article history: Received 16 May 2013 Accepted 9 July 2013 Available online 29 July 2013 Keywords: D. Weldedjoint C. Modification ofsprayingfused E. Fatigue strength abstract Modification ofsprayingfused (MSF) of plasma arc as heat source was used to improve the fatigue per- formance ofwelded joint, which both fundamentally reduced stress concentration at weld toe and achieved metallurgical bond between sprayingfused coating and welding. The lowtransformation tem- perature alloypowder was applied to the method of MSF. After spraying fusion, especially sprayingfusedjointbylowtransformationtemperaturealloy powder, the distribution of residual stress is more difficult to be obtained. Finite element (FE) simulation as an important tool was used to determine the stress field and temperature field ofsprayingfused joint. Simulated results show that as-welded joint and weldedjointsprayingfusedby conventional nickel base alloypowder (Conventional-joint) present tensile stress. The stress ofweldedjointsprayingfusedbylowtransformationtemperaturealloypowder (LTT-joint) is compressive stress. Fatigue test results indicated that under the condition of 2  10 6 cycles, the fatigue strength of as-welded joint is 135 MPa, while that of Conventional-joint and LTT-joint is 218 MPa and 235 MPa, respectively. The fatigue strength of Conventional-joint increases by 61.48%, and fatigue strength of LTT-joint increases by 74.07%. Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved. 1. Introduction Welded joints are widely used in the engineering structure. Most ofwelded joints work under the condition of cyclic loading. Therefore, fatigue failure is one of the most common and danger- ous failure ways [1]. There exists serious structural stress concen- tration and a certain number of welding defects at weld toe. Welding defects is equal to the natural fatigue crack source, greatly reducing the time of crack initiation. So as to the fatigue life is mainly decided by the crack propagation. Consequently, it is very meaningful to take technologic treatment to reduce the stress con- centration and eliminate defects for the improvement offatigue strength [2–5]. The existing technique of prolonging mainly consists of TIG- dressing [6,7], welding toe grinding and ultrasonic peening treat- ment (UPT) [8–11]. TIG-dressing and welding toe grinding are important industrial techniques, but they are faced with the risk of reducing the strength ofweldedjoint due to the change of cross-sectional area. UPT [12–17] mainly rely on the presence of compressive residual stresses to improve fatigue lives ofwelded joint. Tensile residual stresses in welding area weaken the tensile load capacity. As a result, the fatigue life is shortened. For that reason we attempt to reduce or at least eliminate these harmful stresses. By means of periodic impacts or peens with frequencies above 20 kHz, UPT creates a deformation in tensile zone on the sur- face of weld and/or weld toe. Tensile stresses can be relieved by this action. Moreover, cracks and voids are closed, and weld toe geometry is modified. Compressive residual stress is induced by the regional plastic deformation as well. However, UPT also has some shortcomings. For example, the residual compressive stress will be released with the increase of stress level. Thus, UPT will be hard to maintain its ability to resist fatigue. In general, UPT are not suitable for structures operating at applied stress ratios R > 0.5 or maximum applied stresses above around 80% yield. In this study, modification ofsprayingfused (MSF) as a new method to improve the fatigue strength of cruciform weldedjoint was proposed. Deng et al. also studied this new technique [18]. MSF fundamentally reduces stress concentration at weld toe. Meanwhile, metallurgical bond between sprayingfused coating and weld can make weldedjoint bear much larger plus alternating load. In this paper, the heat source of MSF is a plasma flame, which has excellent performance. In addition, the lowtransformationtemperaturealloypowder was applied to the method of MSF, which will make the sprayingfused coating form compressive residual stress. Such, the MSF with lowtransformation tempera- ture alloypowder both reduce the stress concentration and form the residual compression stress, which will further improve the fa- tigue performance ofwelded joint. 0261-3069/$ - see front matter Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.matdes.2013.07.029 ⇑ Corresponding author. Tel./fax: +86 0431 85094687. E-mail address: zhaoxiaohui@jlu.edu.cn (X. Zhao). Materials and Design 53 (2014) 490–496 Contents lists available at ScienceDirect Materials and Design journal homepage: www.elsevier.com/locate/matdes 2. Experiments 2.1. Testing specimens and method The basic principle of MSF (see Fig. 1) is to spray fusing a met- allurgical bonding layer of high curvature appearance on the sur- face of the whole welded joint, which can greatly reduce the stress concentration ofweldedjoint [18]. In order to reduce defects and enhance the bonding strength, this paper use plasma flame in- stead of oxygen acetylene flame to MSF. Specimens were cruciform welded joints (carbon-dioxide arc welding with substrate material of low-carbon steel Q235B and welding material of H08Mn2Si), whose geometrical characteristics are shown in Fig. 2. Table 1 is mechanical properties of Q235B steel. Sprayingfused materials was nickel-base alloypowder (Ni65Cr16B3.1Si4.5), also including lowtransformation tempera- ture alloy powder. They have good oxidation resistance, impact toughness, corrosion and heat resistance. 2.2. The choice oflowtransformationtemperaturealloypowder According to actual circumstances, the author assumes that: 800 °C is a plastic transition point, there is no thermal contraction stress under the condition of this temperature. The relationship between M s and stress under the constraint condition is analyzed. Fig. 1. Basic principle of MSF. Fig. 2. Geometric dimensioning of as-welded joint. Table 1 Mechanical properties of Q235B steel. Material Yield strength r s (MPa) Tensile strength r b (MPa) Elongation (d) Elasticity modulus E (GPa) Poisson’s ratio ( m ) Q235B 267.4 435.5 26% 206 0.3 Fig. 3. Development of stress during constrained cooling. Fig. 4. Finite element model of as-welded joint. Fig. 5. Finite element model ofsprayingfused joint. X. Zhao et al. /Materials and Design 53 (2014) 490–496 491 Literature [19] shows that different combinations of Ni and Cr will be of different linear expansion coefficient and strain, etc. Tensile stress ( r tensile ) is formed due to volume shrinkage of material during the cooling process since temperature decreases from 800 °C. Phase transformation stress ( r compressive ) generated due to volume expansion caused by martensite phase transforma- tion when the temperature drops to phase transition point (M s ). The interaction of phase transformation stress ( r compressive ) and previous tensile stress form the final residual stress ( r ): r ¼ r tensile þ r compressive ð1Þ r tensile ¼ E Á De l ð2Þ r compressive ¼ E Á De p ð3Þ where E is the average elastic modulus; D e l is strain shrinkage caused by heat bilges cold shrink of material; D e p is expansion strain caused by martensite phase transformation. From formula (1)–(3), Stress–Temperature curves of five kinds oflowtransformationtemperaturealloypowder can be gained (see Fig. 3). Fig. 3 shows that the largest residual compressive stress was obtained at room temperature when M s is about 200 °C. Based on the above analysis, the phase transition point ofalloypowder should be controlled at around 200 °C. The chemical composition ofalloy system (wt%) is: C < 0.12, Cr = 5–8, Ni = 8, Mn = 1.0, Si = 0.8, Mo = 0.5, Ti = 0.07, Cu = 0.3, S = 0.005, P = 0.005, B = 0.8, also including a moderate amount of rare earth elements, allow- ance for Fe. 2.3. Method and parameters of MSF In this paper, plasma flame is used to modification processing. Before spraying, the surface ofweldedjoint was pre-grinded, which not only eliminates oxide in the surface ofwelded joint, but also maintains certain roughness of the surface, thus improv- ing the adhesion strength between coating and welding. The pro- cess of modification of plasma spurt spraying only need to change horizontal swing of spray jet into circular orbit rotation on the basis ofspraying process, which can be completed by robot of controlling nozzle line. The basic parameters for this test is: turning radius of spray jet of 200 mm; swing Angle of about 5°; oscillation frequency of 1 times/s. The feeding switch is off after forming overall shape, the non-transferred arc of plasma arc is used to adjust edge position of spray welding layer at a small scale. Meanwhile, for different alloy powder, the process parameters can be appropriate adjusted. 2.4. The temperature field and stress field ofsprayingfusedjoint The residual stress of cruciform weldedjoint is difficult to be measured accurately by the traditional measuring method because of its complicated structure. After spraying fusion, especially sprayingfusedjointbylowtransformationtemperaturealloy pow- der, the distribution of residual stress is more difficult to be ob- tained. Finite element (FE) simulation has become an important tool for the prediction of residual stresses ofwelded joint. 3D FE modeling of a single pass cruciform weldedjoint and sprayingfusedjoint are set up by ANSYS. According to geometric dimen- sioning of as-welded joint (see Fig. 2), the modeling of as-welded joint is showed in Fig. 4. According to geometric dimensioning ofsprayingfused coating, the modeling ofsprayingfusedjoint is showed in Fig. 5. Fig. 5 shows that the minimum thickness of coat- ing is 0.5 mm, and knuckle radius ( q )is20mm. The method of coupled thermal stress analysis is adopted. Firstly a transient thermal analysis is finished to predict the tem- perature filed of the whole joint and sprayingfused coating. Then the results of the thermal analysis are applied as a thermal body load in a transient structural analysis to calculate stresses. In this process, the technique of ‘element birth and death’ is used for mod- eling of the welding and coating material. Heat source is applied in the form of heat flux density. Heat flux density is calculated according to the actual parame- ters ofsprayingfused coating and welding. The total quantity of heat source can be obtained through the current (I), voltage (U) and thermal efficiency ( g ). The total quantity divided by the total volume is per unit volume of heat, and Heat flux density (q)is calculated according to the formula of Q v = q à t(q), t is the time corresponding to a loading step during the process of simulation. The size of heat flux density is q = 7.90  10 10 w/m 3 s in this simulation. Table 2 Hardness and Young’s modulus of Conventional-joint at three different positions. Conventional-joint Coating Transition zone Welding Hardness H (GPa) 6.417 3.829 1.944 Young’s modulus E (GPa) 225.221 182.55 196.209 Fig. 7. Temperature field ofsprayingfused coating. Fig. 6. The bonding condition of Conventional-joint between coating and welding. 492 X. Zhao et al. /Materials and Design 53 (2014) 490–496 3. Fatigue test Specimens were divided into three groups: as-welded joint, weldedjointsprayingfusedby conventional nickel base alloypowder (Conventional-joint) and weldedjointsprayingfusedbylowtransformationtemperaturealloypowder (LTT-joint). All fatigue tests were carried out under constant amplitude load, in term of stretch bending combined loads, with stress ratio (R) of 0.1. Tests were carried out on 100 kN HF Fatigue Testing Machine, whose static load error for full measuring range is ±0.2%. Besides, error of the dynamic load is ±2%. 4. Results and discussion 4.1. The mechanical property ofsprayingfusedjoint The purpose of MSF is to improve fatigue strength ofwelded joint. The adhesion strength between coating and welding which greatly affects the fatigue strength is very important. Fig. 6 showed the bonding condition between coating and welding of Conven- tional-joint. From Fig. 6 we can see that coating and welding achieved perfectly metallurgy bonding. There is no impurities and air hole exit. Therefore, it is difficult for fatigue crack source to form at the transition zone. Fig. 8. The cooling process ofsprayingfused coating. Fig. 9. The extraction path of residual stress. X. Zhao et al. /Materials and Design 53 (2014) 490–496 493 The Hardness and Young’s modulus of Conventional-joint at three different positions (coating, transition zone and welding) based on nano-indentation were shown in Table 2. From Table 2, the value of Hardness and Young’s modulus gradually increases from welding to coating. In addition, the Hardness and Young’s modulus of LTT-joint are basic the same with Conventional-joint. 4.2. Simulated results Because of the distribution of residual stress is decided by tem- perature profile, we first study the distribution oftemperature field. Stress field of LTT-joint is simulated by changing the linear expansion coefficient of Conventional-joint, therefore, Conven- tional-joint and LTT-joint share a temperature field. Fig. 7 is temperature profile ofsprayingfused coating at some time. In the process of simulation, cooling process oftemperature field is very important. The sample finally is used at room temper- ature, so the temperatureof the spray welding layer should be reduced to room temperature. Fig. 8 is cooling process ofsprayingfused coating. Fig. 8a shows that the distribution of isotherm is uniform during the cooling process ofsprayingfused coating. The larger temperature gradient does not exist at each location. Fig. 8b is the final cooling effect, and the whole temperatureofsprayingfused coating is around 22 °C, basically achieving the de- sired cooling effect. Residual tensile stress ofweldedjoint will reduce fatigue strength of joint. Thus, the distribution of residual stress of spray- ing fusedjoint is very important. The extraction path of residual stress was shown in Fig. 9. Because the loading direction offatigue specimen is the direction of X axis, we only analyzed stress distri- bution of direction of X axis. Fig. 10 is stress cloud chart of as- welded joint. Fig. 11 is stress cloud chart ofweldedjointsprayingfusedby conventional nickel base alloypowder (Conventional- joint). Fig. 12 is stress cloud chart ofweldedjointsprayingfusedbylowtransformationtemperaturealloypowder (LTT-joint). Fig. 13 is stress distribution of X-direction of As-welded joint, Con- ventional-joint and LTT-joint along path P1 in the same coordinate system, and Fig. 14 is Stress distribution of X-direction of three kinds of joints along path P2 in the same coordinate system. From Fig. 13 we can see that as-welded joints present large ten- sile stress in this direction. Maximum value of tensile stress reached 294 MPa. The stress of Conventional-joint is also tensile stress at weld toe. But compared to as-welded joint, the value of tensile stress is smaller, only is 107 MPa. The stress of LTT-joint is compressive stress at weld toe of coating, and the maximum value of compressive stress reached À126 MPa. The residual com- pressive stress in the direction offatigue loading will significantly increase the fatigue life ofwelded joint. From Fig. 14 we can see that as-welded joint and Conventional-joint present tensile stress Fig. 10. Stress cloud chart of as-welded joint. Fig. 11. Stress cloud chart of Conventional-joint. Fig. 12. Stress cloud chart of LTT-joint. Fig. 13. Stress distribution of X-direction of three kinds of joints along path P1. 494 X. Zhao et al. /Materials and Design 53 (2014) 490–496 of 300 MPa in this direction. The stress of LTT-joint is compressive stress. Therefore, the fatigue crack sources of as-welded joint and most of Conventional-joints formed at weld toe for residual tensile stress. However, fatigue crack source of LTT-joint has been transferred to the base metal for residual compressive stress. 4.3. S–N curve In order to further verify the accuracy of simulation results and residual compressive stress effects onfatigue life ofwelded joint, high cycle fatigue tests were done. For high cycle fatigue, the max- imum stress is always lower than yield strength of material. When stress ratio is 0.1, stress range is also certain lower than yield strength. Fatigue life corresponding to different stress range can be gained byfatigue tests. S–N curves (The authorized surviving fraction is 50%) ofwelded joints (as-welded joint, Conventional- joint, LTT-joint) are shown in Fig. 15. Fatigue strength of three kinds ofjoint corresponding to 2  10 6 cycles were shown in Table 3. Through calculation, it is clear that under the condition of 2  10 6 cycles, the fatigue strength of as-welded joint is 135 MPa, while that of Conventional-joint and LTT-joint is 218 MPa and 235 MPa, respectively. The fatigue strength of Conventional-joint increases by 61.48%, and fatigue strength of LTT-joint increases by 74.07%. 4.4. Fracture analysis For as-welded joint, the fatigue crack sources of fractured spec- imens are all at weld toe. For weldedjoint with the spray coating (Conventional-joint, LTT-joint), the locations offatigue crack source of fractured specimens are shown in Fig. 16. From Fig. 16 we can see that, most of crack sources of Conventional-joints ap- pear in the weld toe between spray coating and base metal (see Fig. 16a), and most of crack sources of LTT-joints appear in the base metal (see Fig. 16b). Meanwhile, there is also a small part of Conventional-joint fractured in the base metal and a small part of LTT-joint fractured in the weld toe between spray coating and base metal. 5. Conclusions Modification ofsprayingfused (MSF) of plasma arc as heat source both fundamentally reduced stress concentration at weld toe and achieved metallurgical bond between sprayingfused coat- ing and welding. This method also can form compressive residual stress on the surface ofweldedjointby means ofalloypowder with low phase transition point, which will significantly improve the fatigue performance ofwelded joint. The application of numer- ical simulation method plays an important role for understanding mechanism and characteristics of MSF. Fatigue testing results showed that under the condition of 2  10 6 cycles, the fatigue Table 3 Fatigue strength of three kinds of joints (R = 0.1). Type ofjointFatigue strength (2  10 6 ) D r (MPa) Slope (m) Increase degree (%) As-welded joint 135 6.23 – Conventional-joint 218 9.88 61.48 LTT-joint 235 20.28 74.07 Fig. 16. The locations offatigue crack source. Fig. 15. S–N curves of three kinds of joints. Fig. 14. Stress distribution of X-direction of three kinds of joints along path P2. X. Zhao et al. /Materials and Design 53 (2014) 490–496 495 strength of as-welded joint is 135 MPa, while that of Conventional- joint and LTT-joint is 218 MPa and 235 MPa, respectively. The fati- gue strength of Conventional-joint increases by 61.48%, and fatigue strength of LTT-joint increases by 74.07%. In conclusion, modification of sprayingfused (MSF) of plasma arc as heat source will have more widely practical value by adjusting the different alloy powder. Study of MSF will play a promoting role for the development of life-extending technology ofwelded joint. Acknowledgements This work was financially supported by the National Science Foundation of Tianjin through Grant No. 50805102. The authors wish to thank Professor D.P. Wang and associate professor C.Y. Deng for their experimental supports. References [1] Lixing Huo. Fracture behavior and assessment ofwelded structure. Beijing: China Machine Press; 2000 . [2] Dahle T. Design fatigue strength of TIG-dressed welded joints in high-strength steels subjected to spectrum loading. Int J Fatigue 1998;20(9):677–81 . [3] Janosch JJ, Koneczny H, Debiez S, Statnikov ES, Tryufyakov VI, Mikheev PP. Weld World 1996;37:72–83 . [4] Statnikov ES, Muktepavel VO, Blomqvist A. Weld World 2002;46(3/4):20–32. [5] Manteghi S, Maddox SJ. Methods for fatigue life improvement ofwelded joints in medium and high strength steels. IIW Doc. XIII-2006-04; 2004. [6] Ramalho Armando L, Ferreira José AM, Branco Carlos AGM. Fatigue behaviour of T welded joints rehabilitated by tungsten inert gas and plasma dressing. Mater Des 2011;32(10):4705–13 . [7] Morisada Yoshiaki, Fujii Hidetoshi, Inagaki Fuminori. Development of high frequency tungsten inert gas welding method. Mater Des 2012;44:12–6 . [8] Yildirim Halid Can, Marquis Gary B. Fatigue strength improvement factors for high strength steel welded joints treated by high frequency mechanical impact. Int J Fatigue 2012;44:168–76 . [9] Mori T, Shimanuki H, Tanaka M. Effect of UIT onfatigue strength of web-gusset welded joints considering service condition of steel structures. Weld World 2012;56(9–10):141–9 . [10] Abdullah Amir, Malaki Massoud, Eskandari Ahmad. Strength enhancement of the welded structures by ultrasonic peening. Mater Des 2012;38:7–18 . [11] Yang Xinjun, Zhou Jianxin, Ling Xiang. Study on plastic damage of AISI 304 stainless steel induced by ultrasonic impact treatment. Mater Des 2011;36:477–81 . [12] Huther I, Lieurade HP, Souissi R, Nussbaumer A, Chabrolin B, Janosch JJ. Analysis of results on improved welded joints. Weld World 1996;37(5):242–66 . [13] Branco CM, Maddox SJ, Infante V, Gomes EC. Fatigue performance of TIG and plasma welds in thin sections. Int J Fatigue 1999;22(6):589–602 . [14] Kirkhope KJ, Bell R, Caron L, Basu RI, Ma K-T. Weld detail fatigue life improvement techniques. Part 2: application to ship structures. Mar Struct 1999;12:477–96 . [15] Kirkhope KJ, Bell R, Caron L, Basu RI, Ma K-T. Weld detail fatigue life improvement techniques. Part 1: review. Mar Struct 1999;12:447–74 . [16] Marquis G. Failure modes and fatigue strength of improved HSS welds. Eng Fract Mech 2010;77(11):2051–62 . [17] Roy S, Fisher JW, Yen BT. Fatigue resistance ofwelded details enhanced by ultrasonic impact treatment (UIT). Int J Fatigue 2003;25:1239–47 . [18] Zhao Xiaohui, Dongpo Wang, Deng Caiyan. J Mater Process Technol 2011;211:2039–44 . [19] Wenxian Wang. Study onlowtransformationtemperature welding electrode and its application to improve fatigue performance ofwelded joints. PhD thesis, Tianjin University, Tianjin; 2002. 496 X. Zhao et al. /Materials and Design 53 (2014) 490–496 . as -welded joint, welded joint spraying fused by conventional nickel base alloy powder (Conventional -joint) and welded joint spraying fused by low transformation temperature alloy powder (LTT -joint) . All. of welded joint spraying fused by conventional nickel base alloy powder (Conventional- joint) . Fig. 12 is stress cloud chart of welded joint spraying fused by low transformation temperature alloy. by conventional nickel base alloy powder (Conventional -joint) present tensile stress. The stress of welded joint spraying fused by low transformation temperature alloy powder (LTT -joint) is compressive