Effects of maximum hardness in the HAZ on root crack initiation in the JIS-y (oblique-Y groove) cracking test are discussed. The values of critical hydrogen concentration Hc at the location of root crack initiation at 100°C on the way of cooling, were calculated, assuming uniform diffusion of hydrogen through weld zone. The value of log Hc decreased linearly with an increase of Hmax, depending on the intensity of restraint. The correlation between llog Hc and Hmax was found to be far poorer than that between log He and Pcm, Ito and Bessyo carbon equivalent, which is adequate for low carbon content steels. An analysis by the cracking parameter PHA, proposed by the author considering local accumulation of hydrogen at the site of crack initiation, revealed that log Hc vs. Hmax relation is severely affected by the C% of test steels, thus a wide scatter in Hc values. For the same value of Hmax, a lower carbon content is more favourable against root cracking. The estimation agrees well with the JIS-y test result. The critical values of Hmax for crack-free welding without preheating, have been PHA-estimated depending on HD values. As an application, the PHA-estimated critical values of Hmax agree satisfactorily with the Harasawa-Hart's observed values in the H-type self-restrained root cracking tests.
In this study, the weld cold cracking sensitivity of TMCP steels with continuous casting process was investigated. Especially, the effect of center segregation on the weld cold cracking in the through-thickness direction was closely studied. The results in this study were summarised as follows. (1) In the rolling direction, the TMCP steels had high resistance to the weld colcd cracking. (2) The TMCP steels are produced through the continuous casting and thermo-mechanical controlled rolling process. The segregation of P and Mn sometimes occured at the center of plate. It appeared that the center segregation, which was heated above the transformation temperature, result in the hardened region. (3) Hydrogen delayed cracking initiated in the hardened regions after welding under the stress (especially the through-thickness direction stress). It was caused by the hardened structure and stress concentration at inclusions in this region. (4) The fracture surfaces in the regions were terrace feature and contained hydrogen induced intergranular and quasi-cleavage fractures with inclusions.
In this paper, a series of theoretical analyses are performed for restraint stress-strain produced in the weld metal of the oblique Y-groove weld cracking test specimen, varying the kind of steel, heat input, preheating temperature and the base plate thickness. Then, the mechanical characteristics of this specimen are clarified. Moreover, the dynamical measure in the cold cracking parameters (Pw and PHA) are examined from the dynamical view point. The main results are as follows: 1) Even though the size of the specimen is specified, the distribution and magnitude of restraint stress-strain along the slit vary if the kind of steel, preheating temperature, plate thickness, etc. vary. Whereas, if the plate thickness is the same, the restraint intensity is constant irrespective of these changes. 2) Analyzing the data of approximately 500 cracking tests, the validity and practicability of restraint stress-strain to use as a dynamical measure are demonstrated. 3) Critical restraint stress-strains for cold cracking of three kinds of high strength steel are estimated. 4) A new selecting method of the critical preheating temperature Ti* for prevention of cold cracking based on restraint stress-strain is introduced. 5) For three kinds of high strength steel, simple expressions for Ti* are developed.
In welding of high strength steel, toe crack and/or weld metal crack often occurred though the restraint was relatively small. To reproduce such cracks, the strain-concentrated type multi-pass welding crack test was devised. Since they were classified into cold cracks due to diffusible hydrogen, the most effective countermeasurement for preventing weld cracking was to control the preheat and interpass temperature in the second side welding, but the critical preheat and interpass temperature increased in the welded joint whose welding strain was large. It was also clarified that the welding strain produced by the second side welding concentrated at the gorge of weld metal and toe of the first side welded zone as hot strain in the temperature range of blue embrittlement. To investigate the effect of hot strain on the cold crack sensitivity of toe, tensile tests for synthetic heat affected zone (HAZ) were examined. Though the notch tensile strength without diffusible hydrogen at room temperature inrceased due to work hardening by hot strain, the notch tensile strength with diffusible hydrogen decreased due to the combination of hot strain embrittlement and hydrogen embrittlement. As a result the cold crack sensitivity of toe where hot strain concentrated in the temperature range of blue embrittlement was higher than that estimated by Ceq and PCM through JIS oblique y groove welding crack test.
The importance of convective heat transfer in weld pool has been widely recognized in relation to the penetration shapes of base metal. The main causes that associate streaming in molten puddle are the electromagnetic force, surface tension, airodynamic drag force and buoyancy force, but quantitative effect of each factor has not been well analysed. The author have, therefore, conducted numerical analyses of the fluid flow induced by surface tension in order to clarify the independent contribution of thermocapillary effect to the liquid motion in weld pool. A simplified model was employed where the liquid was contained in a hemispherical basin of constant wall temperature and the liquid center was kept different but constant temperature. Under the assumption of laminar flow in a basin, the behaviours of stream lines, velocity and temperature distributions in a pool were obtained. The result gained were as follows: 1) The temperature distributions on the free surface as well as the heat flux distributions on the basin wall became quite different from those of conductive heat transfer as the convection grew stronger in the basin. 2) The effect of natural convection in ordinary weld pool whose size is in the order of centimeter (10°cm) could be reasonably neglected.
A flight process of fused metal in arc spraying is divided into four stages, that is, fusion stage from wire, flying stage in arc plasma, flying stage in the air, and adhesion stage on substrate. The fused metal temperatures are estimated at tow stages on the flight process. One of them, Tc, is estimated at the adhesion stage with a silicone oil calorimeter, and the other, Ti, at the fusion stage by the calculation of heat input. The following results are obtainable from the steimations, (1) Tc and Ti depend on a wire feed speed parameter v+/(v++v-), and the Tc versus v+/(v++v-) curve shows a similar shape to the Ti curve, where v+ is the positive electrode side and v- negative. (2) The temperature Tc at the adhesion stage is higher than temperature Ti at the fusion stage through out the experimental values of v+/(v++v-). (3) The heating effect of the fused metal in the arc plasma is thought to be greater than the cooling effect of the fused metal in the air. This accounts for the higher temperature attained at the adhesion stage than the temperature at the fusion stage.
In fabricating nuclear power plant components resultant safety is of most importance, and it is desirable to use highly reliable welding methods. The authors studied the electron beam weldability of austenitic stainless steel from the viewpoint of joint properties of welded joints and the welding practice. The results obtained are summarized as follows: (1) Basic welded joint properties are good but the tensile strength of welded joint is slightly lower than that of the base metal. (2) The high cycle fatigue strength of welded joint without reinforcement is similar to that of the base metal. The fatigue strength of welded joint with reinforcement, however, is lower than that of base metal with the fatigue strength reduction coefficient of 1.6 to 1.8 at 105 to 106 cycles. (3) The groove gap of up to 0.4 mm at plate thickness of 20 mm, up to 0.5 mm at plate thickness of 50 mm, and up to 0.6 mm at plate thickness of 80 mm enabled welding to be carried out without concavities forming. (4) The treatment of start and end is performed with no problems and the discharge holes can be repaired by applying electron beam welding several times.
On the electron beam (EB) welding process for thick plates, crater treatment to avoid weld defects at the end part of weld bead is one of the most important and very difficult technical problem to overcome. Generally the crater treatment in EB welding is done by the gradual decrease of the beam current, but in a plate of the thickness more than 50 mm, weld defects such as porosities tend to remain at the crater treatment area, where the electron beam does not penetrate through out the plate thickness. This type of defects appears mostly at the tip of the penetration, because, in this area evaporated metal gas cannot be discharged due to the insufficient flow of molten metal. Authors studied various methods of crater treatment of EB welding process for 100mm thick steel plates for pressure vessels. It was confirmed through the studies that porosities can be prevented by the application of double electron beam oscillations (circular and linear direction) with cirtain frequencies and amplitudes.
Local vacuum electron beam welding has been developed to apply it to pipe welding from the inside of a pipe. The gun of this machine with the maximum output of 15 kW (30 kV×500 mA) is able to weld pipes which are 267 mm in outer diameter and 6 m in length. In order to establish pipe welding by using local vacuum electron beam welding, we investigated the method of local vacuum seal, the control of the penetration bead formation, and the problem about the electron gun discharge. The results obtained are as follows: (1) Local vacuum seal mechanism by combining the air expansion tube and 0 rings can keep the vacuum pressure lower than 3×10-3 Torr. (2) The uniform and stable penetration beads are formed by controlling piercing beam current in welding carbon steel pipes of 9.3-30.9 mm in thickness in flat and vertical positions. (3) It is effective for preventing weld defects caused by the electron gun discharge to provide a highspeed turn off-turn on circuit on high voltage secondary side of the power supply.
In the fracture toughness tests of welds, it is often found that fracture toughness values measured with reference to cleavage fracture initiation show a great amount of scatter even if testing conditions are much the same. Such scattering characteristics are also recognized in tests of macroscopically homogeneous steel, and in this case, the scatter is assumed to be caused by a statistical nature due to microstructures of the material. In the tests of weld specimens with macroscopic heterogeneity along the crack front, for example cross-bond-type notched specimens, however, the scatter in cleavage resistances may be affected not only by heterogeneity in microstructures but also by macroscopic heterogeneity in toughness. In the present paper, using a few types of bending COD specimens with macroscopic heterogeneity along the crack front, the effect of macroscopic heterogeneity on the scatter in fracture toughness values have been investigated. Moreover, a probabilistic model of the scatter of fracture toughness of welds is proposed, based on the following assumptions; (1) the lowest toughness along the crack front controls the cleavage fracture toughness, that is, the weakest link model is available, and (2) the size of the embrittled region along the crack front controls the scatter in fracture toughness values. Comparing the distribution of fracture toughness values calculated by the present probabilistic model with that obtained from the tests, applicability of the model proposed in the present study is clarified.
Fracture toughness of welds has been usually evaluated, based on COD tests developed for homogeneous material. However, since mechanical heterogeneity caused inevitably in welds considerably influences crack opening behaviour, critical COD calculated according to the conventional method does not necessarily represent an appropriate fracture toughness parameter of material in the vicinity of crack tip. In the present study, at first, "Local COD criterion" has been proposed newly as an appropriate fracture criterion of welds with mechanical heterogeneity. In order to clarify the availability of the new criterion, bending COD tests and tensile tests of notched-welded plate have been conducted, using specimens extracted from undermatched and evenmatched weld joints of HT80 steel. FEM analyses have been also conducted in order to characterize the deformation behaviours of these specimens. From the experiments and analyses, in the welded joints with notch existing along fusion line, Local COD δL defined newly in our study is proved to control strains in the vicinity of crack tip, and :applicability of Local COD δL as an appropriate fracture characteristic parameter is clarified. Moreover, it is suggested that fracture performance of welded joints with mechanical heterogeneity should be evaluated, based on Local COD criterion proposed in the presnnt study.
Critical COD values of weld metal obtained by standard bending COD test are often scattered widely. One of the causes of the scattering will be local embrittlement-zones in weld metal which are inevitable in multi-pass welding. In the present study, statistic investigations of the critical COD δc obtained by both bending tests and tensile tests are made. Standard bending COD specimens and notched-welded tensile specimens are cut from weldments of a steel for low temperature use, 22 mm thick. Weld metal of each specimen includes artificial embrittlement zones of different sizes ranging between 4 mm and 22 mm in depth. Tensile tests are made by a 600 tons testing machine with a coned disc spring. It is demonstrated that the standard bending COD test is much sensitive to existence of local embrittle-ment-zone in weld metal than notched tensile test: Cumulative frequency of the δc values in bending test is almost the same independent to the depth of embrittlement zone. On the other hand, the δc values in tensile test are increased with decrease of the depth, and in the tensile specimens with an embrittlement zone of 4-5 mm in depth, final fracture is occurred at net stress level as high as yield stress of the material, although in some specimens the first fracture is observed as pop-in at lower COD level. Statistic study reveals that the δc values of the tensile specimens with an embrittlement zone ranging 4 mm-18 mm in depth are 1.5-2 times the δc values of the bending specimens.
This work deals with the initiation of hydrogen-induced disbonding which occurs at overlay welds of austenitic stainless steel from view of metallurgy. An experiment was done about specimens exposed to elevated temperature under high pressure hydrogen atmosphere. Hydrogen-induced disbonding was initiated on the inside of carbide layer, which was formed at the interface of overlay and base metal by post weld heat treatment, with the duplex structure of austenite and martensite and/or the austenite structure with 0-10% ferrite respectively. The range of chemical compositions showing these structures corresponded to 20-28% Ni equivalent (Ni*eq). At these structures by post weld heat treatment, the formation of carbide layer exhibited itself at the austenite grain boundary of the transition zone and accompanied the formation of depletion zone of chromium, α' martensite and stacking fault at that region. Hydrogen-induced disbonding seems to be caused by the connection between the above metallurgical factors and hydrogen accumulation to the interface of overlay and base metal.
In the previous paper, the relation between the initiation of hydrogen-induced disbonding and the structure at the interface was investigated. As a result, hydrogen-induced disbonding occurred at the structure with 20-28% Ni* equivalent. In this work, on the basis of Ni equivalent, the conditions necessary for prevention of hydrogen-induced disbonding were discussed in view of welding materials, the dilution ratio and the condition of post weld heat treatment. Hydrogen-induced disbonding did not occur at the structure with Ni* equivalent less than 20% or more than 28%. On the other hand, hydrogen-induced disbonding can be prevented or lightened by controlling dilution ratio or the time of post weld heat treatment respectively.
Combined effect of phosphorus, chromium and molybdenum on the reheat cracking sensitivity was investigated using synthetic Cr-Mo steels melted in our laboratory. The cracking sensitivity was evaluated in the terms of critical restraint stress obtained by the modified implant method applied for the reheat cracking. The results are summarized as follows. (1) In the case of a constant phosphorus content of 0.01%, the combined effect of chromium and molybdenum (0 to 3%Cr, 0.5 to 1.4%Mo) was shown by menas of the contour lines of the critical restraint stress in the Cr-Mo content diagram (Fig. 3). (2) In the case of five sets of Cr-Mo contents, the effect of phosphorus in a content range of 0.004 to 0.1% was examined. There exists a critical phosphorus content Pcrit, from which the cracking sensitivity increases with the increase of phosphorus content. Pcrit values are: above 0.02%, 0.01%, 0.008%, 0.014% and below 0.01% for 0%Cr-0.5%Mo, 0.5%Cr-0.5%Mo, 1%Cr-0.5%Mo, 1.3%Cr-0.5%Mo and 2%Cr-l% Mo steels, respectively. (3) Phosphorus segregation at the prior-austenite grain boundary was estimated by a grain boundary etching method. Phosphorus has segregated there already in as-welded condition, and the degree of segregation corresponds to the cracking sensitivity. (4) Grain boundary concentration of phosphorus becomes the maximum by tempering at about 500°C, which corresponds to the temperature of crack initiation.
The transformation behavior of the ferritic weld metal in 9%Ni steel was investigated using a dilatation technique. (1) Along dendrite interfaces of the ferritic weld metal, segregations of silicon, manganese and nickel are formed. (2) The Ac transformation temperatures of the ferritic weld metal are lowered mainly by the increased nickel addition (11%Ni) to the weld metal. (3) On account of the nickel segregation along the dendrite interfaces, the Ac transformation temperatures of the interfaces are further lowered. (4) In the reheating process of the ferritic weld metal, it commences the true Ac1 (austenite) transformation from the dendrite interfaces. (5) The optimum tempering temperature of the ferritic weld metal is in the range from the true Ac1 temperature to about 600 degree C. The micro-struc-ture of the tempered ferritic weld metal comprises ferrite, retained austenite and cementite precipitating along the dendrite interfaces. (6) At reheating temperatures higher than about 600 degree C, the martensite transformation begins to occur from the dendrite interfaces during cooling. At a temperature of about 700 degree C (Ac3), the whole structure becomes martensite. (7) At reheating temperatures higher than about 1100 degree C, the bainite transformation followed by the martensite transformation is performed during cooling. (8) It is attributable to the manganese segregation which lowers the carbon activity that the cementite precipitation during tempering is concentrated at the dendrite interfaces. (9) Although the toughness of the ferritic weld metal is deteriorated, the sites of the cementite precipitation can be dispersed by increasing silicon content in the weld metal.
The grain growth of austenite in a heat-affected-zone (HAZ) has been generally described by the following equation, Dn-D0n=k0t exp(-Q/RT), where D is average grain diameter, D0 is initial grain size, n and k0 are material constants, t is annealing time, Q is experimental activation energy for grain growth, R is gas constant, and T is absolute temperature. The above material constants are obtained by the isothermal annealing experiment on heating processes, but thermal cycle of welding is composed of heating and cooling process. Therefore, it is necessary to consider the grain growth in cooling process. In this paper, the material constants in heating and cooling processes were examined by isothermal annealing experiment, respectively. Then, the grain growth in the HAZ was separated into heating and cooling processes, and the relation between grain growth and welded thermal cycle were examined. Main conclusions obtained in this report are summarized as follows; (1) The difference of grain growth between heating and cooling processes in the HAZ did not show. The material constants obtained, (a)In the case of n=2 k0=2.408×104mm2/s, Q=51100cal/g·atom (b)In the case of n=4 k0=2.208×108mm4/s, Q=89200cal/g·atom (2)The average grain size in the HAZ could be estimated by using n=2 and material constants of (1).
The present paper is concerned with the residual stresses and deformations caused by a circumferential and/or longitudinal welding of thin-walled cylindrical shell of mean radius α, wall thickness h arid length l. Based upon experimental data by GMA welding process and analytical results by the previous papers, controlling parameters or combined effect of heat input and size of cylinder on the residual stresses and deformations are discussed. In the circumferential-welded cylinders, the residual stresses σθ/σy, σx/σy in circumferential and longitudinal directions and deflections ω/aεy at a distance x/√ah from the weld centerline are determined by the two parameters, Heat input parameter Hθ=Q/cρh√ah Size parameter βθ=l/√ah , where σy, εy, c and ρ are yield stress, yield strain, specific heat and density of the material respectively, and Q is net heat input per unit weld length. When the size parameter βθ exceeds a critical value, (βθ)cr, depending upon the heat input parameter Hθ, the residual stresses and deformations are determined uniquely by the heat input parameter Hθ. In the longitudinal-welded cylinders, the residual stresses σx/σy, σθ/σy, and deflections √h/aω/lεy at an angle θ from the weld centerline on the mid-cross seetion are decided by the parameters, Heat input parameter Hx=Q/2cρπah Size parameter βx=(l/a)√h/a In the range that the size parameter βx is larger than its critical value, (βx)rc≅3, only one controlling parameter is the heat input parameter Hx. When the size parameters exceed their critical value, the residual stresses and deflections are expressed as functions of the heat input parameters. The calculated values are approximately agreed with the experimental results.
Consideration is conducted on parameters controlling deformations of thick cylindrical vessels due to narrow-gap multi-pass welding of longitudinal joints. The change in inner diameter and the deformations in the vicinity of welds caused by longitudinal welding is measured for actual cylindrical vessels, 50-200 mm thick, 900-3000 mm in inner diameter. Experiments for thick plate specimens under the same narrow-gap welding process are also made. In the multi-pass welding of thick plate, the deformation occurs after the manner of angular change as the root part of groove hardly moves. Therefore the amount of angular distortion is obtained by Δgw/h using the measured value of shrinkage Δgw at upper side of groove and plate thickness h. The above characteristics in deformation of multi-pass welds of both cylindrical vessel and plate specimen result in usefullness of the application of elastic theory based on dislocation model for estimating the deformation of cross-section of vessel due to longitudinal welds. According to the dislocation theory, change in inner diameter ΔDi is proportional to the inner diameter Di and is given by the equation ΔDi=-Δgw/h·1/2π(1-1/2πsinθ)·Di(0°≤θ≤180°) where the sign of ΔDi is defined as increase in Di is given as positive, and θ is the angle from the weld joint. The applicability of the above equation based on dislocation theory is confirmed by the present experiments. In the above equation, the term Δgw/h of angular distortion of thick cylindrical-vessel has a constant value regardless of inner diameter and plate thickness when plate thickness is larger than 50 mm. Accordingly, it is presumed that the change in inner diameter is controlled by only one parameter of inner diameter of vessel.
A new sampling inspection system of NDT has been proposed to secure the quality of welds of steel buildings. In this system, the quality of welds is evaluated based on reliability-analysis in which probabilistic parameters such as detectability of NDT, sampling rate are considered. In the present new system, the whole processes of welding works are divided into a certain number of steps, and then, the reliability of every step is calculated by using the inspection data of that step. And the inspection plan is modified at any next step if necessary so that the average reliability of all welds will conform to satisfy the reliability requirement which was already given by the structural designer. In the present paper, the new system is termed "Step Inspection System". The availability of Step Inspection System to secure the required reliability of welds was confirmed by applying actually to the 40-storied and 27-storied steel buildings.
The toughness of SUS 316L MIG weld metal was investigated in relation to the oxygen content, pass sequence, δ ferrite content and PWHT conditions, and also the effect of δ ferrite content on the hot crack susceptibility of the weld metal was discussed. (1) The toughness is greatly improved by decreasing oxygen content in the weld metal to about 50 ppm. The value of VE-269 reaches to as high as 15.2 kgf·m. (2) Welds with low oxygen contents can be attained by using rare-earths-bearing welding wires in a pure argon shield. (3) For preventing reheat embrittleness, decreasing δ ferrite content to about 1% and using narrow-gap MIG arc welding with 1 pass/1 layer are effective. (4) When the δ ferrite content is too high, the toughness of weld metal is deteriorated. This is because cracks propagate along the δ ferrite where precipitates of M23C6 produced by multiple welding thermal cycles are existing. (5) Decreasing δ ferrite content to about 1% is not so harmful to the hot crack resistivity of the weld metal so far as the contents of silicon, phosphorus and sulfur are sufficiently low. (6) Since PWHT deteriorates the weld metal toughness, it is desirable to be avoided. In the case that PWHT is required, the one at a relatively low temperature (about 600°C) is recommended.
The formability of butt welded stainless steel sheets was investigated through Erichsen cupping and deep drawing tests. In the case of austenitic stainless steel SUS304, even TIG Welded sheet with wide fused zone showed the same degree of formability as that of base material, taking care against lubricant. In the case of ferritic stainless steel SUS430, the EB welded sheet with narrow fused zone was better than that of TIG welded, sheet. The formability of welded sheet of modified SUS430 with lower C and with ferrite-forming element was nearly equal that of base material.
The corrosion protective properties of the sprayed aluminium coating in hot water were investigated by the immersion test. Main results of the experiment are summarized as follow; (1) The sprayed aluminium coating protects the steel substrate with the electrochemical effect in hot water at 60°C. (2) The potential of the sprayed aluminium coating shifts slightly towards the noble potential with decreasing water temperature. (3) When mild steel with sprayed aluminium coating and stainless steel are immersed in, an electrolyte and in electrical contact, the dissolution rate of the sprayed aluminium coating is accelerated and its protective life becomes shortened. The galvanic current between the sprayed aluminium coating and stainless steel decreases with decreasing water temperature. Copper ion accelerates.the dissolution rate of aluminium in galvanic corrosion.
To study the effect of Al in 50 kgf/mm2 class C-Mn-Nb steels on weld metal microstructure and toughness, single-pass, bead-in-groove tandem submerged-arc weld metals have been made at 49 kJ/cm. The work was based on a 1.5%Mn-0.5%Mo wire, with two kinds of highly basic agglomerated fluxes and one low basicity, flux. Semikilled, Si-killed and Si-killed, Al treated plates of 25 mm thickness were employed. The preliminary findings presented in the report showed that even small changes in weld metal Si and Al levels (e.g. 0.02 to 0.03% Al) could produce drastic deterioration in Charpy toughness (up to 100 deg C shift in 3.5 kgf⋅m temperature) when using some basic fluxes. Furthermore, the results indicated that achievement of high weld metal toughness with some basic fluxes might be difficult in high dilution welding of Si-killed, Al-treated steels having more than about 0.035-0.045% Al. The effect of Al on microstructure and toughness was found to be independent of its source (wire or plate).
To investigate effects of nitrogen contents in wires and plates, carbon contents in wires, flux chemistry and welding condition on nitrogen absorption by weld metal and slag during submerged-arc welding, singlepass beads have been deposited on Si-Mn steels of 20 mm thickness by using the combination of four kinds of agglomerated and fused fluxes, having different basicity, with seven kinds of Si-Mn welding wires, having different carbon and nitrogen contents. The results showed that only the molten weld pool produced by highly basic fused flux absorbed nitrogen from the arc atmosphere in addition to wire and plate when welding at high speed, whilst at low welding speed no nitrogen absorption occurred from the arc atmosphere independently of flux type or flux basicity and hence nitrogen content in weld metal was almost the same as all amounts of nitrogen which came from wire and plate. Nine kinds of Si-Mn steels and two kinds of CaO-MgO-Al2O3-SiO2 slags have been melted by a tungsten arc in a copper cruicble within Ar-N2 atmosphere in order to examine effects of nitrogen contents in arc atmosphere and steels, steel oxygen content, slag basicity and melting time on nitrogen contents in arc-melted metals and slags. Nitrogen dissolves into molten metal from arc atmosphere and then it is partially transferred into slag according to a dissociation reaction of silica. This nitrogen transfer decreases with increase in slag basicity, ilicon and oxygen contents in metals because of a reduction of silica dissociation. Molten slag without the coexistence of molten metal does not absorb nitrogen from arc atmosphree. Nitrogen dissolution into molten metal is accelerated in oxygen rich metal to some extent when no metal coexists with slag.
Investigations into relationship between Ti content and toughness of 50-60 kgf/mm2 tensile strength submerged-arc weld metals containing different amounts of nitrogen to obtain excellent, tough weld metals and also effects of oxygen and Al content on the optimum Ti content have been studied. An addition of 0.04 to 0.08% Ti into weld metals from welding wires gives excellent toughness to the weld metal containing nitrogen of 25 to 80 ppm. The optimum Ti content increases with nitrogen content in weld metal so that improvement of weld metal toughness by Ti addition to weld metal is not considered to be based on only the precipitation of TiN particles as an acicular ferrite nucleation sites during transformation from austenite. The contribution of Ti to improvement of weld metal toughness stems from fine acicular ferrite formation accompanied with reduction in free nitrogen. Decrease in weld metal oxygen content or addition of Al into weld metal leads to shift the optimum Ti content to lower levels and to decrease precipitated Ti and increase precipitated N. It is therefore considered that with decreasing oxygen in weld metal Ti fixes nitrogen more effectively because the major precipitated Ti consists of oxide and nitride and that Al combines with oxygen in preference to Ti, allowing ready combination of Ti with N, because Al addition produced little AIN precipitation. As a result, addition of small amounts of Ti to weld metal, linking with decrease in weld metal oxygen and nitrogen content may well be key point of improvement of weld metal toughness.