An experimental investigation of weld pool oscillations was carried out for stationary pulsed TIG arc weld pools using analogue transistor power source. Oscillations of molten metal were electrically detected with the aid of accelerometer and observed by high speed photography. Arc force was also evaluated from pressure distribution of TIG arc on the water cooled copper anode.
Results obtained are summarized as follows;
(1) It is possible to detect resonance frequency of weld pool using the measurement system of mechanical vibration, which is consisted of accelerometer and specimen attached on the beam.
(2) Larger weld pools are found out to have lower resonance frequency in stationary pulsed TIG welding.
(3)In the frequency range lower than the fundamental natural frequency of weld pool, amplitudes of weld pool oscillations can be evaluated through measuring the acceleration of vibration of the beam.
(4) Arc force is almost proportional to the square of current applied. In pulsed TIG welding, molten pool is drived to oscillate by fluctuation of arc.force due to plasma stream.

The high speed one-side welding of thin stainless steel using TIG welding method is applied to the pipe production widely. In this paper, the bead formation and th e welding stability under above mentioned process are investigated.
The stability fo one-side welding of butt joint is pertinent to the undercut bead formation. Highly stable one-side welding without undercut is established by inhibiting metal flow toward the freezing side. For this purpose, it is more effective to hold forward torch angle and to use knife edge electrode dispersing the arc force. Especially, using knife edge electrode with forward torch angle, stable one-side welding of 0.2-0.4 mm thick plate under 15 m/min is possible easily.

A possibility of a back bead width control system have been proposed to stabilize the process of one side FCB SAW in which a stationary copper backing plate covered with a flux layer of several milimeter thickness is used as backing.
When a stable welding process is maintained under an adequate welding condition, a signal voltage between the backing copper plate and the plate being welded have beeen detected. The polarity and phase of the detected voltage having a few volt level is well corresponding to those of the arc voltage in case of AC current welding.
In a stable FCB SAW process, the back side bead is mostly formed by a key-hole action of the arc. It is assumed for the reason that the arc penetrates through' the welding groove and the flux layer, then it contacts the copper plate and is reflected to the welding plate. The correlation between the detected voltage level and the back side bead width is recognized.
The similar phenomenon is observed in a key-hole welding by using a transfer type plasma arc welding.
In order to maintain the back bead width constant the detecting voltage is compared with a reference voltage, and then the welding current is controlled by the difference of the above voltages.

Behavior of oxide film on void surface of the diffusion-welded interface has been investigated with Auger electron spectroscopy, SEM and EPMA. Results obtained are summarized as follows.
1.The oxide film of aluminum on the void surface did not disappear during welding. However, the oxide film of copper, iron and SUS 304 stainless steel disappeared, forming lumps of oxide on the void surface owing to spheroidizing of the oxide film, wheres the lumps of oxide was not observed on the free surface heated in a vacuum. On the other hand, the oxide film of titanium disappeared without forming lumps of oxide.
2. Disappearance of the oxide film of copper, iron, SUS 304 stainless steel and titanium was owing to diffusion of the oxide film into the base metal.

The effect of an insert metal on the diffusion bonding of alumina ceramics to carbon steel has been invsetigated with particular reference to the relaxation of thermal stress which is set up in the cooling process from bonding temperature owing to the difference in thermal expansion between the alumina and steel. The insert metal used was a combination of a titanium foil 20μm thick and a copper foil 0.05-4mm thick. All joints obtained were fractured in the alumina on tensile test. However, the fracture strength increased with the increase in the thickness of the copper foil, approaching a saturated value (30MPa) for foil thickness of 2-4mm. In the alumina ceramics cracks were observed in the as-bonded state. The crack decreased in length with the increase of the thickness of the copper foil and could not be observed when the foil thickness was more than 2mm. According to a stress analysis using finite element method (FEM), the crack was initiated at the position where the thermal stress took a maximam value, and the direction of the crack was nearly normal to that of the maximam stress. The analysis using FEM indicated also that the thermal stress decreased with increasing the thickness of the copper foil because the plastic flow of the copper foil relaxed the stress. Consequently, the formation of the crack in the alumina ceramics is attributable to the thermal stress caused by the difference in the thermal expansion between the alumina and steel. The crack and thermal stress are considered to cause the change in the fracture strength of the joint with the thickness of copper foil. A combination of the titanium foil and a nickel foil 2mm thick was also applied as an insert metal. However, the fracture strength of the joint was considerably low.

In this work, the trapping effect of inclusions on hydrogen diffusion was investigated and it was considered how this trapping effect affected on hydrogen embrittlement.
The results in this work were as follows;
(1) At room temperature diffusion coefficients were lowered and nondiffusible hydrogen contents were increased with increasing inclusions. Therefore, it was aquired that inclusions acted as both reversible and irreversible trapping sites in steels.
(2) It was appeared that nondiffusible hydrogen was trapped in inclusions and microvoids produced at inclusions at room temperature.
(3) It was appeared by measurement of binding energy between trapping sites and hydrogen that reversible trapping effect of inclusions on hydrogen diffusion was caused by dislocation fields developed around inclusions during oil quenching of steels.
(4) It was concluded by considering the model of hydrogen embrittlement proposed by Beachem that the index of hydrogen embrittlement was relative to concentration of hydrogen at the crack tip. And it was appeaered that inclusions lowered the concentration of hydrogen, thus elevated K_TH.
(5) K_TH values were elevated by Ce treatment, while diffusivity of hydrogen was not affected by Ce treatment. The reason of this result was caused by the upward shift of critical line in Beachem model because of higher K_IC of Ce treated steel.

In this work, the trapping effect of inclusions on hydrogen diffusion was quantitatively evaluated and numerical method were applied to analyze local hydrogen concentration at the crack tip. The results obtained in this report are summarized as follows;
(1) The validity of hydrogen diffusion model being based on effective hydrogen concentration was confirmed in the case of varying micro structure, plastic strain and inclusion contents.
(2) The trap parameter, by which micro structure, plastic strain and inclusion contents were quantitatively evaluated, was proposed. And this trap parameter was applied to hydrogen diffusion equation based upon the concept of effective hydrogen concentration.
(3) Hydrogen concentration at the crack tip was analyzed by using the finite element: method based upon above hydrogen diffusion equation. From this result it was appeared that hydrogen concentration at the crack tip was slowed down and the maximum content of hydrogen accumulation was decreased with increasing inclusion contents. Furthermore, it was shown by this analysis that with increase of inclusions critical hydrogen concentration for crack initiation was lowered inspit of the incuvation time being inceased. This result agree with the previous conclusion obtained from the model of hydrogen embrittlement.
(4) Also FEM analysis was carried out in the case of considering stress induced hydrogen diffusion. And similarly above effect of inclusions was shown.

Comparison between Implant test, RRC and side-slotted JIS-y (Tekken) tests has been performed using the authors' new cracking parameter P_HA which was introduced previously based on JIS-y test data.
New formulae for predicting critical rupture stresses in 8-7 mm D Implant test have been developed by multiple regression analysis of 102 test data on HT50 to HT80 low-alloy high strength steels of low carbon contents of 0.04 to 0.18%. The formulae proved to be more accurate than conventional formulae thus far proposed.
The 8-7 Implant critical rupture stresses of the tested HT50 to HT80 steels were compared with the critical crack initiation stresses in RRC and JIS-y tests predicted by PHA analysis. Equivalence of critical stress values between 8-7 Implant and RRC or side-slotted, JIS-y tests has been found to exist only in a limited stress range of 340 to 540 N/mm², provided the test steel, electrode hydrogen content, and the cooling time are kept the same, respectively, in both the tests. In the higher stress range exceeding 540 N/mm² in HT50 steels, the 8-7, Implant critical stress condition which leads to microvoid coalescence type fracture, does never seem to be realized in RRC and JIS-y tests.
The estimated relationship between critical preheating temperature and Pcm obtained via Implant test seems to be practically identical to that via RRC or side-slotted JIS-y test, only in the case of critical stress level of approximately 451 N/mm². Taking the critical stress to be equivalent to the yield stress of base metal seems to be reasonable only in the case of the said stress level and practical welded joint of mild restraint.

Toe crack and/or weld metal crack often occurred in double side welded joint. It has been appeared in previous work that the factors causing these cracks had been due to both effects of hot strain caused by angular distortion at blue embrittlement temperature range and diffusible hydrogen from 2nd side weld.
In this study, to investigate the effect fo hot strain on cold crack sensitivity of toe and weld metal, using 14 commercial steels and weld metals, notch tensile strength was measured for hydrogen charged synthetic heat cycle specimen pre-strained at blue embrittlement temperature range. And local strain for fracture was evaluated, converting the notch tensile strength into local strain by Neuber's method. Furthermore, considering both results of above tests, a formula to estimate the critical preheat-interpass temperature in multipass welding using Ceq, weld strain, sol. nitrogen, atmosphere and restraint intensity was proposed.

Microcracks on the fillet joint between the lowest course shell plate and the annular plate are often detected by M.T. after hydrostatic test. These are delayed crackings due to hydrogen. No one has been studying these cracks in the oil storaged tank. Authors have studied the simulation of these cracks by R.T.T. apparatus and three point bending test. Results are as follows.
1) Microcracks could be simulated by R.T.T. apparatus and three point bending test.
2) Hydrogen distribution in the fillet weld joint is not uniformly. Hydrogen gathers beneath the final pass.
3) Most of microcracks were observed at the regions where hydrogen gathered and hardness increased.
4) Microcrack susceptibility increased by grinder dressing because it makes subsurface area harden.

After simulation of microcrackings by three point bending test, effect of some parameters on the microcrackings are studied. Results are as follows.
1) Without hydrogen charge, microcrackings were not detected by M.T.. In this case, those were observed by microcsope, and crack distribution was submitted to logarithmic normal distribution. Mean crack propagation rate in the observation area was 7.1×10^-3 mm/hr.
2) With hydrogen charge, crack susceptibility by M.T. increased as current density increased, or as deflection increased. Crack number and crack ratio increased with current density. Test coupon with grinder finish contained more microcrackings than that of as welded.
3) With hydrogen charge in sea water, microcrackings initiated on the surface of test coupon.

Weld hot crack in Fe-36%Ni alloy which is used as elements of'membrane-type container for liquefied gas was studied. One of the characteristics of the hot crack is that it occurs in weld metal reheated by the following welding pass. Its dynamic behavior in Cross-bead tensile hot cracking test was analyzed with direct observation technique (MISO technique). Regression analysis was done on chemical composition in relation to total crack length in Cross-bead tensile hot cracking test conducted for 24 different materials. Materials insusceptible to the hot crack were produced on practical production line, and used for practical fabrication of model tanks. Main conclusions are as follows: 1) Weld metal reheated to the range of 600°C to 1000°C by the following welding pass is susceptible to reheated hot crack. 2) Microstructural and fractographic features suggest that the reheated hot crack is ductility-dip crack. 3) S, O, N and Al are crack-promoting elements, while P shows a slight trend of crack-promoting. 4) Model tanks fabricated with materials insusceptible to the hot crack showed favorable result in inspections.

Simulated hot ductility test was done to establish the correlation with reheated weld hot cracking test of Fe-36%Ni alloy and to reveal the mechanism of reheated hot crack. Cross lines were drawn with a sharp razor on specimen surface electropolished to study the effects of grain boundary sliding, grain boundary migration, recrystallization and so on. Main conclusions are as follows: 1) There was ductility trough between about 600°C and 1000°C in reheated weld metal, and minimum ductility was located arround 900°C. These behaviors agreed well with those in the actual reheated weld hot cracking test. 2) There was good correlation between brittleness temperature range or minimum ductility evaluated by hot ductility test and total crack length in Cross-head tensile hot cracking test. 3) Grain boundary sliding together with grain boundary precipitates and/or grain boundary serration seems to cause the ductility trough and thus reheated ductility-dip crack by cavity formation process. The grain boundary precipitates contained Al and S. 4) Recovery of dutcility at about 1100°C was associated with active grain boundary migration and recrystallization.

The effect of hardness and grain size of steels and their welds on fracture stress and sensitivity of Liquid Metal Embrittlement Cracking in molten Zinc (LMEC) was investigated. The experiment was conducted using commercial weldable structural steels and Sustained Load Test (SLT) was conducted in molten Zinc with heating range of 430°C-470°C.
The results obtained are as follows.
(1) The LMEC behavior of steels and welds in molten Zinc is of thermal activated process, and fractures are intergranular (IG) type which corresponds to actual LMEC in hot dip galvanizing.
(2) Estimation of fracture stress of steels and welds in molten Zinc is obtained by measuring Vicker's hardness and grain size, and these estimation values are correspond with test results.
(3) The sensitivity of LMEC is affected by hardness of steels and welds.

The liquid metal embrittlement cracking (LMEC) of heat affected zone (HAZ) of steel in hot dip galvanizing was investigated. The experiment was conducted using synthetic HAZ of JIS G 3444 STK55 and Sustained Load Test (SLT) at 370°C-470°C heating range in liquid (molten) and solid Zinc.
The results obtained are as follows.
(1) LMEC behavior is of thermal activated process and LMEC sensitivity is affected by hardness of HAZ.
(2) The coarse grain region in HAZ showed the highest LMEC sensitivity.
(3) The interfacial energy of steel is reduced by the existence of Zinc.
(4) The reaction rate process of LMEC of steel is controlled by grain boundary diffusion of Zinc.
(5) The nucleation and the propagation of LMEC in steel are likely to have a close relation to the reduction of interfacial energy due to the grain boundary diffusion of Zinc.

Influence of microstructure in the Heat-Affected-Zone on the reheat cracking susceptibility was studied.
The reheat cracking tests on H.T. 80 steel were carried out by means of ring type specimens and three point bending test. Influence of microstructure of synthetic H.A.Z. at high temperature on the reheat treatment process were investigated by high temperature hardness measurements, X-ray analysis and scanning electron microscopic observations. The results are summarized as follows. 1) The reheat cracking's critical stress of Martensite structure specimens (Water Quenched) were generally double than that of Bainite and Ferrite structure specimens (Air Cooled). But, the reheat cracking susceptibility of A.C. specimens were little by the difference of forms in microstructure. 2) The reheat cracking susceptibility were estimated by the relation of the deformability of a material and the precipitate of alloying carbides. Accordingly, alloying carbides of A.C. specimens precipitated to M₃C type carbides during cooling process of synthetic H.A.Z., but that of W.Q. specimens precipitated to M₃C type carbides as well as A.C specimens and M₂C type carbides during reheat treatment process. W.Q. specimens of the lower reheat' cracking susceptibility are affected mainly by the precipitate of M₂C type carbides in the matrix. 3) It is seemed that the precipitation of M₂C and M₃C type carbides during reheat treatment process from W.Q. specimens decreased the alloying elements concentration in the matrix, and showed a good deformability of a material. As a consequence, the reheat cracking susceptibility of W.Q. specimens were lower than that of A.C. specimens.

The dissolution of TiN in a matrix during synthetic weld thermal cycling was investigated in detail and at the same times, the relationship between the amount of soluble Ti, soluble N, and TiN in a thermal cycled specimen and were investigated quantitatively in this study. The results obtained are as follows;
(1) The amounts of insoluble Ti (soluble Ti) and insoluble N (soluble N) are not affected by Al content.
(2) It is considered that TiN particles dissolve depending upon a solbility product of TiN. (3) And the temperature dependence is denoted by the next equation,
log (wt%Ti)(wt%N)=-7600/T-0.011
where, 0.01≤Ti≤0.03 wt%, 0.002≤N≤0.008 wt% and 1200≤T≤1410°C. In case of specimens containing a few amount of Ti like less 0.006 wt%, the another smaller solubility product of TiN was obtained and the value was 4.9×10^-6 at 1350°C. (4) As the precipitation of TiN was not observed during the cooling process of the thermal cycle, the amounts of soluble Ti and soluble N depend upon the solubility product of TiN at the maximum temperature of the thermal cycle. (5) Transition temperature, _vT_rs of synthetic weld heat affected zone depends strongly upon the forms of Ti and N, that is shown the following equation,
_vT_rs=0.229 Sol.Ti+0.0991 Sol.N+0.047 Insol.Ti-75.1
It is very significance for toughness improvmeent to decrease soluble N content. (6) Fraction of fine grained ferrite-pearlite structures increases with the increasing Ti and N content in specimens keeping the stoichiometric ratio of Ti/N=3.42. (7) However, the notch toughness deteriorates a little with the increasing Ti and N content in spite of a constant amount in soluble Ti and soluble N. (8) It is made clear from observation of TEM and SEM that TiN particles of above 0.2 μm in size increase with the increasing Ti and N content in case of Ti/N=3.42, and also TiN particles of above 0.5μm in size were observed on an unit facet. It is considered that these large TiN particles cause the initiation and the propagation of brittle cracks and therefore deteriorates the notch toughness.

In the gas shield arc welding with SUS304 austenitic stainless steel base metal and, JISY308L type electrod, the effect of the welding condition on the oxygen content and microstructure of weld metal has been investigated. Next, the effect of the oxygen on the mechanical properties, mainly on the fracture toughness, were studied. The main results are summarized as follows;
1) The oxygen content of weld metal is increased with dicreasing the welding current (from 200A to 400A) and with increasing the arc voltage (from 20V to 30V). At the constant welding atmosphere, there is a linear correlation between the oxygen content and a parameter of A/V which is established by the authors, where A is welding current and V is arc voltage. The oxygen content decreases with increasing the parametre A/V.
2) The most of the oxygen exists in the γ region as the spherical oxide of Mn-Silicate. The number of oxide per unit volumu is related to the welding heat input. The number decreases with increasing the heat input.
3) Charpy impact values and Jin values decrease with increasing the oxide volume fraction.
4) The relationship between Charpy impact value and oxide volume fraction is represented by linear line on the log-log cordinate. At the constant oxide voulme fraction, the distribution (number and size) of oxide has little effect to Charpy impact value.

This investigation was carried out to clarify the cause of cracks which are often detected in the 11/4 Cr- 0.5Mo steel chemical reactor vessel in use. Furthermore, the prevention of such cracks and rapid evaluation of creep embrittlement were also examined.
Firstly, appearance of cracks detected was investigated by the light microscope and SEM. The cracks were identified as the typical intergranular fracture initiated and propagated in the coarse grained HAZ adjacent to the fusion line of weldment. Grain boundary segregation of impurities such as Sn, Sb and S was revealed by Auger electron spectroscopy.
Secondary, several kinds of test including creep rupture test and high temperature tention test with slow extention rate technique (SERT) were conducted with the synthetic weld HAZ specimens directly sampled from the materials in which cracks were detected. Results of the creep rupture test indicated the creep embrittlernent behavior of specimens clearly by the low creep ductility and the small notch rupture strength ratio. Creep rupture strength of notched specimen coincided with the stress applied to the part of the reactor vessel where cracks were detected. As a result, it was verified that cracks detected were due to the creep embrittlement phenomenon.
Thirdly, it was confirmed that reheat treatment (900°C) was available to prevent the creep embrittlement of coarse grained weld HAZ.
Finally, SERT test using notched specimen was usefull for the accelerated evaluation of creep embrittlement by taking notice of fracture mode. In addition to high temperature atmosphere, such technique is also available for reappearance of damage caused by high temperature and high pressure hydrogen surroundings.

The objectives of the present paper is to clarify the effect of yield ratio of material used on structural performance and to consider the possiblity of making low-yield-ratio HT80-class steel. Mechanical analyses on deformations of typical structural elements have been conducted to determine the appropriate condition of yield ratio of materials. Based on the analysis of plastic deformation of dual-phased structure steel, consideration has been carried out of the making process of low yield ratio HT80 by using the same kind of steel as the conventional QT-HT80.
The decrease of yield ratio is effective procedure for increasing deformability of structural element. In HT80 steel, controlled rolling and on line accelerated cooling process followed by quenching from α+ γ phases is a proper treatment for obtaining low-yield-ratio steel with dual-phased structure in which martensite structure includes a certain fraction of lower strength ferrite.

Metallurgical consideration and elementary performance tests have been conducted of low yield ratio HT80 steels made by quenching process from α+γ phases termed Q'. Nine kinds of steels subjected to various heat treatments and their welds are used for the present experiments such as tensile tests, Charpy test and bending COD test.
By using TMCP process followed by Q'-quenching and tempering, about 85% of yield ratio of HT80 steel can be realized even if the chemical composition is same as one of conventional HT80 steel. It is made clear from metallurgical observations that the reason why that kind of steel becomes low yield ratio is due to the existence of pure ferrite in the matrix of tempered martensite. The CR-Q'-T steel proposed in the present paper has higher deformability and fracture toughness than those of conventional HT80 steel.

The transient thermal stresses of 13 Cr cast steels (SCS5), austenitic stainless steels (SUS304) and their weld metals are investigated by round bar specimens which both ends are fixed and subjected to thermal prestraining cycles simulating weld heat. The experimental results in comparison with SUS304 steels indicate that the effects of transformation expansion on residual stress distributions are remarkable, and the effects of multi-thermal prestraining cycles on residual stress distribution are negligible small in SCS5 cast steels. These results are compared with longitudinal residual stresses of butt welded joints, and it is shown that the both-ends bar testing model is available to simply evaluate residual stress characteristics. It is observed that the residual stress of SCS5 welded joint is about -100 MPa at weld metal and the residual stress of SCS5 welded joint prepared with an austenitic weld metal is about +380 MPa. Three dimensional residual stress distributions of 13 Cr cast steel weldments prepared with an austenitic weld metal are measured by Ueda-Fukuda's method. It is observed that the maximum residual stress is +450 MPa at midthickness, which is nearly equal to 0.2% proof stress of weld metal.

The authors developed the measuring theory of three dimensional residual stresses, in which inherent strains, the source of residual stresses, are used as parameters, and demonstrated L_z method and nL_y, method. In these methods, inherent strains are estimated using relaxed strains measured at many points. Necessarily, cutting is required many times. In the cutting process, additional inherent strains may be produced by the effect of cutting heat. Therefore, it is desirable to decrease as much as possible the number of cutting, which is also very important to improve the accuracy of measurement and to shorten the time required for experiments.
In this paper, in order to accurately estimate three dimensional residual stresses produced in a region where residual stresses are uniform along the welded line, a measuring theory and its method using a thin plate cut out perpendicular to the welded line are presented.
The main results are as follows:
(1) Using relaxed strains measured before and after the cutting of T-specimen, effective inherent strain, the source of residual stresses, can be estimated being divided into cross-sectional inherent strain components, {ε^*_y, ε^*_z, γ^*_yz}^T, and one along the welded line, {ε^*_x}.
(2) Giving the estimated effective inherent strains to the original three dimensional object, residual stresses are calculated. By this method, three dimensional welding residual stresses at an arbitrary position including the inside of the object can be estimated.
(3) The reliability of this theory was confirmed by numerical experiments.
(4) The number of cutting has greatly been decreased. Accordingly, the time required for experiments can be greatly shortened and experimental expenses cut down.
(5) This method is applicable to measurement of three dimensional residual stresses produced by any causes as well as welding, if the residual stresses are considered to be uniform in one direction.

Considerations have been conducted fundamentally on stresses·strains controlling parameter of cracked plate joined with two dissimilar materials, especially of which yield strengths are different each other, that is σ^I_Y, σ^I_Y. Attention mainly is paid to the joined plate with a crack of which tip, exists in the vicinity of joining surface of dissimilar materials. FEM analyses are applied for calculation of deformation behaviours under tension of two kinds of cracked plates: L-type specimen in which crack is perpendicular to the joining surface and P-type specimen in which crack is parallel to the ioinine surface.
Mechanical heterogeneity remarkably influences on stresses.strains in the vicinity of the crack tip when plastic deformation spreads beyond the joining surface. On the analogy of linear elastic fracture mechanics theory, the parameter, F, which controls stresses. strains in the vicinity of the crack tip has been proposed. By using two parameters; strength ratio of two materials, Sr(=σ^II_Y/σ^I_Y, and F[=σ_G/σ^I_Y)·√c/d], characteristics of stress.strain distributions of two kinds of cracked plates can be classified.

Initiation of pop-ins becomes a great obstacle in satisfying critical COD requirements for weldments because COD values at pop-ins are often lower than required ones. Though sonic countermeasures for such pop-ins may be possible, it is too difficult to completely avoid them by improvement of materials and welding procedures as the critical COD value is sensitive even to extremely local brittle zones.
Therefore, in the previous report, authors proposed the assessment method for significance of pop-in, which involved approval of pop-in in COD test when it was proved to be arrested in actual structures with reasonable background.
In this report, the proposed assessment method is further investigated in terms of the measuring meth-od and the theoretical approach. The results are as follows:
Dynamic behaviour of pop-in can be measured clearly and conveniently by the newly developed method to which magnetic induction effect is applied. By this method, it was shown that dynamic be-haviour of pop-ins occurred in COD test for an actual welded joint was similar to that for model welded joints, and so proposed assessment method is applicable to COD test on welded joints for practical use.
Finite element analyses made it clear that the 3-point bending COD test was severer than actual structures for propagation and arrest of a small brittle crack when the dynamic change of the K_ID value or the elastic energy release rate was considered. Hence it is pointed out that pop-ins found in the COD test for welded joints with low strength materials can also be arrested in actual structures.

In ASME Code Sec. III, the value of the reference nil-ductility temperature RT_EDT has an important significance to determine the result of the fracture mechanics evaluation. While in the standard both the drop-weight test and Charpy impact test are required to determine the RT_NDT, in practice it is normally determined only by the nil-ductility transition temperature (T_EDT) obtained by the drop-weight test.
The causes of data scatter in T_NDT were investigated to establish appropriate conditions of crack-starter bead welding. Drop-weight tests were carried' out for nuclear vessel steels by changing welding conditions to examine the effects of welding amperage and shapes of welding table on T_NDT. The results show that the preparation of crack-starter bead by small welding amperage should not be allowed, because it makes the measured T_NDT non-conservative, and that it is important to use a welding table which increases the cooling rate of specimen.
Furthermore, the authors proposed methods for estimating T_NDT of nuclear vessel steels by using Charpy transition temperatures.

Since a weld defect may occur at the root of butt welded joint with backing strip, its fatigue strength is lower than conventional butt welded joint. In order to investigate the effect of weld defect on the fatigue strength, it was carried out a series of fatigue tests of welded joint with backing strip containing artificial defect and natural defect. The data of fatigue tests were compared with the calculated fatigue life using fracture mechanics.
As a result, it is confirmed that fracture mechanics is effective method of assessing the fatigue life in the welded joint with backing strip, because the calculated fatigue life is shown to be consistent with the experimental results. The change of propagated crack shape were made clear for the welded joint with backing strip. The S-N diagram of welded joint with backing strip have been proposed and the S-N diagram also offer that this joint is possible to allow up to 1 mm defect.

Stress corrosion crack initiation feature of Type 304 steel in 42%MgCl₂ aq. sol. under mixed mode stress condition was studied. Brittle fracture criteria, Kσ(θ)_max, S_min, G(γ)_max were tried to apply to SCC initiation. As the result, it was made clear that threshold stress field intensity for SCC in mixed mode condition was more adjusted by these criteria than only mode I stress intensity, K_I. In addition, Micro morphology of SCC fracture surface were studied by fractography. Fracture planes in SCC initiation point was mainly constituted by {100} plane and scarcely {111} slip plane which was observed only localized point. This shows that SCC was seemed to be more susceptible to opening mode than shear mode.