Formation of molten metal layer at weld interface was found to be detected in real time by measurement of effective time-integral value of weld voltage as monitoring parameter in upset welding of low carbon steel rod.
Upset weld quality can be assured by formation of uniform molten metal layer at weld interface at early period of welding and of weld interface larger than mother metal in sectional area at final period of welding. The molten metal layer can be formed by melting of the interface due to weld current, and, furthermore, it seems to be connected to effective time-integral value of weld voltage which is related to temperature rise at upset welds. It is experimentally confirmed that the uniform molten metal layer is formed over a certain value of effective time-integral value of weld voltage with no relation to edge angle, welding condition and so on for upset welding process of low carbon steel rod (20mm dia.) with prepared edge.

View full abstract

The sectional area at weld interface of upset welds at final period of welding has been found to be measured in real time by both weld resistance and effective time-integral value of weld voltage for low carbon steel rod.
The current path area is related to weld resistance, current path length, and average resistivity, which is a function of material composition and average temperature at welds, based on Ohm's law. It is clarified that average resistivity at welds can be determined from effective time-integral value of weld voltage. So the sectional area at weld intrefacecan be measured by both weld resistance and effective time-integral value of weld voltage and this method can be adopted as one of monitoring parameter for estimating the welding quality of low carbon steel rod.
According to the results, effective time-integral value of weld voltage and weld resistance are shown to be fundamental monitoring parameters of upset welds with prepared edge. Furthermore, this monitoring system is shown to be effective in assurance of upset weld quailty with no relation to weld condition, edge angle and so on, for low carbon steel rod.

View full abstract

This paper deals with a non-destructive detection method of the generation time, the crack length and the propagation speed of fatigue crack in spot-welded joints under tensile-shear load.
In case of spot-welded joints, fatigue cracks generate usually at the neighborhood of nugget tips of the weld interface, and propagate toward outer surfaces of the sheets. Accordingly, it is very difficult to observe and measure the crack length, consequently the generation time and the propagation speed of crack by using conventional measuring devices.
Authors developed a non-destructive method of fatigue crack, which the generation time, the crack length and the propagation speed of fatigue crack can be quantitatively estimated by utilizing infomation obtained from the strain gages bonded on the electrode indentations of spot welds. Specimens used in this experiment are made of various kinds of thin steel sheets.
Main conclusions obtained in this study are as follows:
(1) The propagation length of fatigue crack is related with both mean strain and strain range of out-put of the strain gage, and the changing rate of strain range is specially useful for fatigue crack detection.
(2) The generation time and the propagation speed of fatigue crack can be estimated easily from relationship between the number of cycles and the crack length, which is established in advance for an individual test piece.

View full abstract

A mathematical model of the weld pool, where the heat input from arc is situated on the molten pool surface, has been developed to discuss the influence of weld pool configuration on the heat flow in arc welding.
The numerical model, based on the heat flow equation in the quasi-stationary state and the surface tensional balance equation for the molten pool, makes it possible to assess the effect of process parameters such as arc pressure and wire deposition rate on the weld penetration.
In the first part of this paper, the finite difference molten pool models have been proposed and the numerical method has been discussed.
In the second part of this paper, the effect of thermal property of material on the weld pool shape has been discussed and the influence of weld pool configuration on the penetration has been made clear.

View full abstract

The optimum relation between welding current and wire melting rate for the high-speed hot wire GTA welding is decided and a new power control system is developed.
The new system has two merits, one is that the power source is simplified by using a part of the arc current as the wire heating current, and the other is that a stable melting of wire is realized by monitoring the voltage drop across the wire extension.
The results in this study are as follows.
(1) The welding speed can be increased by pre-heating the wire. The effect of speed-up is promoted especially in the case that the bead width is enlarged by adding the hot wire deposit.
(2) The optimum relation between welding current and wire melting rate for high speed welding is almost independent of the plate thickness.
(3) In the new control system, the welding speed can be raised up to two times of that obtained with the cold wire GTA welding without troublesome adjustments of the power source.

View full abstract

The spattering in the pulsed GMA welding with a short arc length was investigated and a practical method for suppressing the spatter was presented. An arc length is usually kept short in order to raise the welding speed without a weld flaw. But, in such a case, the short circuit between the wire and the base metal frequently takes place. The spattering was observed with a highspeed framing camera and changes of the gas density around the arc were observed by the shadowgraph method. It was made clear that the spattering was resulted from the short circuit. The spatter was genearted by the rapid Jouleheating at the minimum cross-section of the molten metal.
The practical method for suppressing the spatter was to control the current waveform so that the short circuit may take place during the base duration.
With the inverter controlled power source, the amount of splashed spatter was decreased up to onetenth of that obtained with the conventional pulsed GMA welding power source.

View full abstract

Characteristics and behavior of tungsten arc and feasibility of arc welding of mild steels were investigated in the pressure range from 0 to 5 MPa (gauge). As the cathode, 2% thoriated tungsten of 4 mm in diameter and of 30° in vertex angle was used.
Results obtained are summarized as follows:
(1) Arc voltage increases with increasing pressure. However, V_A+V_K decreases a little with increasing pressure from 0 MPa to 0.5 MPa. It shows about 7.3 V constantly over 0.5 MPa.
(2) Diameter and half width of arc decreases and the current density in centre of arc increases considerably with increasing pressure.
(3) The area of cathode spot decreases and therefore it concentrates at the tip of tungsten with increasing pressure. As the result, the erosion of tungsten tips increases.
(4) In constant welding current, the depth of penetration of mild steel increases with increasing pressure. Therefore, the optimum current for penetration welding of thin mild steel plate decreases with pressure.
(5) Relatively sound penetration welds of mild steel plates of 2 and 3.2 mm in thickness can be obtained under pressure range up to 5 MPa. However, the meandering bead was often obtained by the fluctuated arc over 2 MPa.
(6) Welded joints obtained under pressure range of up to 5 MPa have sufficient tensile strength and Charpy impact value.

View full abstract

A suitable friction welding condition to obtain the joint between pure copper and pure aluminium, which had joint efficiency over 100%, was decided. And the welded joint was pressed in the direction of diameter. Mechanical properties, i.e., tensile strength, bend ductility and hardness, micro-structure and fracture behaviour of the as-welded joint and the pressed joint were investigated.
Reduction R was defined as follows:
R=(d-t)/d×100%
where d: diameter of joint before pressworking,
t: thickness of pressed joint.
Main results are summarized as follows.
(1) Tensile strength of the welded joint increases with an incraese of reduction ratio and tensile fracture of the joint, pressed in the reduction ratio of above 20%, always takes place in aluminium base metal.
(2) In the cases of reduction ratios of 40% and 48%, the joints have sufficient ductility so that they are able to be bent up to 90° without cracking in bend test. It seems that improvement of ductility by pressworking is owing to spherical projection of weld interface from copper side to aluminium side.
(3) When the pressed joint is broken in tensile test, fracture takes place in aluminium at the position adjacent to weld interface, that is, it takes place in the alloyed layer and aluminium base metal and at the boundary between alloyed layer and aluminium.

View full abstract

Characteristics of creep rupture strength and brittleness under steam turbine operation for 2%Cr-1Mo steel plate EB welded joint have been investigated besides that of high temperature strength and stress relief cracking reported already. As the result, high-intermediate pressure outer casing of 380MW steam turbine have been manufactured by welded structure of 21/4Cr-lMo steel plate in stead of Cr-Mo-V cast steel previously used.
Results are summarized as follows;
(1) 10⁵ Hr creep rutpure strength of EB welded joint satisfys with scatter band of base metal and allowable stress indicated by Table UCS23 of ASME SEC. VIII.
(2) FATT of HAZ and weld metal for EB welded joint increase with holding time at 538°C. But it's value saturates more than about 8000 Hr and is lower than that of Cr-Mo-V cast steel.
(3) EB welding have been applied to steel plate welded structure of high-intermediate pressure outer casing for output power 380MW steam turbine. It's maximum thickness EB welded was 110 mm.

View full abstract

The fatigue crack growth behavior near threshold region in a Type 304 stainless steel and its welded joints (GMAW and EBW) was investigated. The behavior in an SM50A steel and its welded joint (GMAW) was also investigated to compare the difference of threshold behavior between austenitic steel and normal pearlite-ferrite steel. Although the effects of residual stress on the fatigue crack growth rate and threshold value were quite significant, that of microstructure of the weld metal was also found to be significant in both welded joints: The growth rates of weld metal specimens were higher than that of parent metal specimens near threshold region when the crack growth rate was plotted against effective stress intensity range.

View full abstract

Fatigue crack propagation under repeated variable amplitude loading in welding tensile residual stress fields was studied experimentally in low propagation rate region (low dK region). Material used was steel of 800 MPa in tensile strength. Specimens were center-cracked-plates with a longitudinal bead on center of specimen.
Major results obtained are summarized as follows.
Fatigue crack propagation rate under repeated variable amplitude loading was estimated from linear accumulation rule for relation between crack propagation rate and dK under constant amplitude loading. Threshold value of dK in the relation, however, had to be changed to the value in appearance.

View full abstract

X-ray stress measurement, calculation of stress concentration factor by FEM, fatigue test, etc. were performed to confirm the effects of weld shape and residual stress on fatigue strength of mild steel non-load-carrying cruciform fillet welded joint.
The results obtained are summarized as follows:
1) A relief of residual stress at the toe by cutting off the large size cruciform fillet welded joint specimen to the small size specimen was about 40%, but the fatigue strengths of both cutting off small size specimen and as welded same size specimen were almost the same. Such a degree of relieving of residual stress by cutting off, therefore, can be considered to have no effect on the fatigue strength.
2) By FEM analysis in the case of axial tension load, the following became clear: Stress did not flow into the rib (cruciform plates) outside fillets, and the stress concentration factor was less than 1.0 constantly at the weld root (tip of slit parallel to the direction of axial load), and on the other hand, was 1.7-2.3 at the weld toe.
3) Using the stress concentration factor in profile of weld toe fatigue fractured pracitcally, and assuming the failure probability, it was found that the fatigue strength of mild steel cruciform fillet welded joint at 2 X 10⁶ cycles could secure about 118MPa {12 kgf/mm²} at failure probability p=5%.
4) It is possible to consider that a fatigue life of peened specimen is improved owing to compressive residual stress and work hardeness (increase of yield point), and decreased owing to increase of surface roughness (5-8 times that of before treatment; Rmax=1518 jem) caused by peening and that the synthetic effect of the former two exceeds the negative effect of the latter, and fatigue ilfe is improved after all.
The effect of compressive residual stress decreased with increasing applied stress, and became uncertain at the applied stress over yield point.
5) By re-peening before or after a initiation of small crack (1, is about 0.5 mm) fatigue strength was improved additionally. We may expect improving fatigue strength in the practical welded structure, therefore, by a practical application of re-peening.

View full abstract

Microcrackings could be simulated in HT60 weld metal by Restrained T type Tensile (RTT) test. They had two types of cracking. One initiated in the weld metal, called type I cracking, and another initiated from the surface of weld metal, called type II cracking. Type II cracking could be detected by M.T., but type I cracking could not be detected by M.T.. Therefore, if type I cracking exists in the fillet weld between the lowest course shell plate and the annular plate in oil storage tank, no inspection method can be applied to detect them. Fatigue test specimen was sampled from RTT test specimen including type I cracking. Various moistening condition of D6216 electrode were applied to initiate microcrackings in the weld metal. Experimental results were as follows.
1) At 350°C, 3.6 ksec drying of D6216, the hydrogen content in HT60 weld metal was 7 cc/100 gFe, and at 80°C, 86.4 ksec moistening of D6216, it was 37 cc/100 gFe.
2) At 350°C, 3.6 ksec drying of D6216, no microcracking was detected in HT60 weld metal.
3) At 350°C, 3.6 ksec drying of D6216, tensile strength of weld metal did not change before and after fatigue test.
4) At 80°C, 86.4 ksec moistening of D6216, number of cycles to fracture increased with decreasing cyclic load.
5) Even if type I cracking existed in the fillet weld between the lowest course shell plate and the annular plate, these might not make a oil storage tank fracture in operation for 50 years.

View full abstract

There are increaseing needs for the fabrication of hydraulic machineries which are used under the severe condition of cavitation. So, the anti-erosion weld deposits are usually overlaid on the position where the cavitation is likely to take place. However, there is large discrepancy that the surfacing weld deposits which has better weldability exhibit less erosion resistance. Increase in the hardness of the deposits is effective to improve its erosion resistance, while it spoils the weld-cracking avoidability of the deposits. In order to prevent the cracking of the deposits, it is necessary to increase its ductility. There should be a strong demand for the development of the surfacing weld deposits of the good combination of the erosion resistance and weldability.
Investigations have been performed on the effects of the contents of carbon, manganese and nickel on the erosion loss and the elongation of the austenitic weld deposits containing 20% chromium and 7%cobalt. Results are as follows;
(1) Erosion loss and its deviation increase as the nickel equivalent increases. It is necessary to limit the maximum of nickel equivalent less than 12.5% in order to assure erosion loss less than cobalt base alloy of low carbon content whose composition system is 0.06%C-19%Cr-9%Ni-15%W.
(2) Elongation behaves likely to erosion loss. It is necessary to limit the minimum of nickel equivalent in order to assure the elongation more than 10%.
(3) The erosion loss Cw (mg) and the elongation El(%) can be respectably estimated by the following experimental formulas. Cw=-64.6+115.60+10.1Mn+18.6Ni and El=-21.4+51.1C+2.5Mn+4.8Ni.
(4) The weld deposits whose chemical composition system is 0.25%C-3%Mn-3%Ni-20%Cr-7%Co have good combination of erosion resistance and weldability.

View full abstract

The purpose of this work is to reveal beneficial effects of rare earth metals (REM) additions on reheat cracking susceptibilities of lCr-0.5Mo and HT80 weldments. Reheat cracking susceptibilities of 1Cr-0.5 Mo steel containing 0 to 0.37 wt% REM and HT80 steel containing 0 to 0.1 wt% REM have been examined at 600°C by a hot tensile test and a small-sized y-slit crack test. Hot ductilities at 600°C of these steels were improved approximately from 10 to 80% with an increase of REM content. Moreover, any reheat cracking did not occur in lCr-0.5Mo steel weldment by adding 0.02 to 0.13 wt% REM. On the other hand, both steels containing more than 0.1 wt% REM were susceptible to liquation cracking. Therefore, the reasonable range of REM content for preventing reheat cracking was approximately from 0.02 to 0.05 wt% in lCr-0.5Mo steel. The reason why REM improved reheat cracking sucseptibilities of these steels was reducing of grain boundary segregation of impurity elements such as P, S etc. on account of scavenging effect of REM.

View full abstract

This study deals with the estimation of occurrence and preventive condition of the Liquid Metal Embrittlement cracking (LMEC) due to hot dip galvanizing using Y-Groove Crack ing Test pieces.
The results obtained are as follows,
(1) Using Y-Groove Cracing Test piece is appropriate for the estimation test of the LMEC in welded steel structures.
(2) The critical condition for LMEC occurrence is indicated by the relation between the intensity of restraint and the maximum hardness in HAZ (Hv max.).
(3) The LMEC preventive condition for welded steel structures to be galvanized are obtained by the Hv max. under the condition that welded joints have the most severerity value of the intensity of restraint in each type of structure. For example, the LMEC preventive condition of weldment in transmission tubular tower structures is assumed that Hv max. keeps less than 260.

View full abstract

Oxygen-free copper plates were welded with copper and copper-titanium alloy electrode wires using argon-nitrogen gas mixture shielding. The effects of the nitrogen percent (N₂%) in the shielding gas and the titanium content (Ti%) of the electrode wire on mechanical properties and annealed structures of the weld metal were examined. The Vickers hardness increases with Ti% of the electrode wire. The yield stress and the tensile strength of the weld metal increase with Ti% of the electrode wire, while the elongation and the reduction of area show maximum at about 0.5%Ti in the wire. Those tensile properties of the weld metal are independent of N₂% in the shielding gas. β' precipitates are observed in some of the weld metals annealed at 573-1173 K for 3.6 ks. The precipitation is the most remarkable after annealing at 700-800 K, and the amount of the precipitates increases with increase of Ti% in the wire and with decrease of N2% in the shielding gas. The Vickers hardness increases with the amount of the precipitates.

View full abstract

Experimental covered electrodes(lime-titania type)were made by selective additions of Mn-nitride or Cr-nitride to the electrode covering. 304LN stainless steel base plates(t=12mm)whose edge preparation was machined to single sided V, 60°included angle were clamped in a jig and were welded together bymulti-pass shielded metal arc welding process.
Nitrogen content of wcld metals was adjusted by changing the content of Mn-nltride or Cr-nitride in the electrode covering.
Effect of nitrogenon the microstructure of weldmetal, δ-ferrite content, mean grain width, hardness and δ-ferrite secondary arm spacing were studied. Impact test for as welded specimen and solution treated specimen have been carried out at temperature range from 77K to room temperature.
Influence of nitrogen on the impact value of wcld metals was mainly discusxsed.
Results obtained are summarized as follows:
(1)δ-ferrite content in the weld metals is influenced largely by the nitrogen content.
(2)Impact value of weld metals increases with increasing of the nitrogen content of the weld metals up to about 1700 ppm after then the impact value falls as the nitrogen further increased.
(3)The strengthening mechanism of weld metals due to addition of nitrogen is estimated that the effect of austenitic structure stabilized and inetrstitial solution hardening. Micro fissuring and precipitation of cluster(nitride)in excess of nitrogen containing weld metals decrease the impact value of weld metals.
(4)Nitrogen content up to about 1700 ppm of weld metal which is introduced by SMAW process gives a good influence on the impact properties at low test temperature under condition of this study.

View full abstract

The first layer overlay weld metal with austenitic/martensitic duplex microstructure provides the overlay welds with better resistance to the hydrogen-induced disbonding than that with the conventional austenitic microstructure containing several percents ferrite as previously reported.
In this report, the notch tensile strength of the bond of the overlay weld the first layer of which is composed of duplex microstructure (the disbonding resistant overlay weld) after absorption of hydrogen was studied and compared to that of the conventional overlay weld.
The main results obtained in this study are as follows:
(1) The notch tensile strength of the bond of the improved overlay weld was higher than that of the conventional overlay weld.
(2) The notch tensile strength of the bond of the conventional overlay weld decreased as decreasing crosshead speed.
(3) The notch tensile strength of the bond of both overlay welds decreased as increasing of PWHT temperature and holding time.
(4) The notch tensile strength of the coarse grain boundary of the conventional overlay weld after PWHT at 973K for 7.2 × 10⁴ S was lower than that of the bond.
It was concluded that the coarse grain in the first layer weld metal adjacent to the bond strongly affected the resistance to hydrogen-induced disbonding of the stainless steel overlay weld.

View full abstract

The first layer overlay weld metal with austenitic/martensitic duplex microstructure provides the overlay welds with better resistance to the hydrogen-induced disbonding than that with the conventional austenitic microstructure containing several percents ferrite as previously reported.
The resistance to the hydrogen-induced disbonding occurring at the bond of the stainless steel overlay weld is affected by the carbide precipitated at the bond region.
Based on this consideration, the properties of the disbonding surface and the carbide precipitation of the disbonding resistant overlay weld were studied and compared with those of the conventional overlay weld.
The main results obtained in this study are as follows:
(1) The disbonding mainly occurred at the bond in the disbonding resistant overlay weld, while it occurred at coarse grain boundaries in the first layer weld metal adjacent to the bond in the conventional overlay weld.
(2) Fine M₂₃C₆ carbides precipitated in the first layer weld metal adjacent to the bond in the disbonding resistant overlay weld.
(3) Large M₂₃C₆ carbides precipitated at the coarse grain boundary in the conventional overlay weld.
(4) The amount of precipitated carbide of the disbonding resistant overlay weld was almost equal to that of the conventional overlay weld.

View full abstract

The first layer ovrelay weld metal with austenitic/martensitic duplex microstructure provides the overlay. welds with better resistance to the hydrogen-induced disbonding than that with the conventional austenitic microstructure containing several percents ferrite as previously reported.
The resistance to the hydrogen-induced disbonding occurring at the bond of the stainless steel overlay weld is strongly affected by the hydrogen concentration at the bond region.
Based on this consideration, the hydrogen distribution in through-thichness direction of the disbond-ing resistant overlay weld was studied and compared with that of the conventional overlay weld.
The main results obtained in this study are as follows.
(1)The hydrogen contents in the first layer overlay weld metals adjacent to the bond of the conventional and disbonding resistant overlay welds were. 25.6 and 16.5 ppm, respectively, at steady state operating condition.
(2) The hydrogen content in the first layer overlay weld metal adjacent to the bond of the conventional overlay weld increased after shut down of operating condition. On the other hand, that of the disbonding resistant overlay welds scarcely increased.
(3) The maximun hydrogen content of the conventional overlay weld described in (2) was 36.5 ppm at 3.24×10⁴ seconds after shut down of operation. That of the disbonding resistant overlay weld was 17.9 ppm at 3.6×10⁴ seconds.
(4) The lower hydrogen content after shut down of operating condition described in (3) provides the overlay weld.with good resistance to the hydrogen-induced disbonding.

View full abstract