1. Introduction

The increase in global shipping traffic has inevitably entailed the construction of larger ships. It is necessary to lighten the structural weight of ships and improve their productivity by using high-strength steel materials. In particular, for large container ships, steel with a thickness of 80–100[mm] is used to ensure sufficient longitudinal strength.

These is referred to as thick steel plates or extremely thick steel plates. Due to the characteristics of the steel, several problems such as brittle fracture can be caused by the thickness of the plates. The thickness effect means that the patterns of failure appear differently depending on the thickness of the steel materials, even for the same material with the same properties. Distribution patterns of stress generated inside the material are classified as planar stress or planar deformation. For thinner plates, the steel is gradually destroyed by plastic deformation due to external forces. In the case of thick steel plates or extremely thick steel plates, deformation is difficult due to the thickness. The disadvantage is that the destruction proceeds rapidly once a certain limit is reached. In addition to these disadvantages, studies have been conducted on fracture toughness through CTOD tests, such as excessive heat input, which also affects brittle fracture.^[1] In addition, research on cooling speed, the effects of tissue, and alloy elements has also been carried out, but does not generally replicate the welding conditions in the field.^[2-7] It is therefore urgent to be able to determine welding integrity because verification is not performed at the locations where the welding takes place.

In this study, behavior patterns that affect the change in the heat input and fracture toughness for the welding joints of thick steel material used in large container ships were investigated.

2. Experiment method

2.1 Mechanical and chemical properties

The steel materials used in this study were EH-40 BCA (Brittle Crack Arrest) with a thickness of 80[mm], which was applied to the upper deck of the container ship using a thermo-mechanical controlled process (TMCP). Table 1 shows the chemical composition, and Table 2 lists the mechanical properties.

The mechanical properties of the applied steel materials needed to be suitable for high-strength welding with a yield strength of 470[MPa]. A flux core wire corresponding to the E81T-C1A6-K2 standard of the American Welding Society (AWS A5.36), was manufactured with a wire diameter of 1.4[mm] and 99.9% CO₂ was used as a protective gas.

2.2 Welding method

The welding improvement shape is shown in Fig. 1 to compare and evaluate the fracture toughness of the weld metal according to the FCAW welding technique for thick steel material.

The steel materials consisted of thick quenched and tempered steel. In the case of the initial bead, the maximum heat input was 55[kJ/cm] it was subsequently maintained at 34.4[kJ/cm]. The following measurements were used: groove angle 30°, root face 0 3[mm], root gap 3 – –8[mm]. The preheating temperature and maximum interlayer temperature were 127[°C] and 170[°C], respectively, following the FCAW technique, and the welding conditions are listed in Table 3.

2.3 Fracture toughness test

If the steel material to be tested is thin, the plastic deformation caused by the yield phenomenon before the crack is formed cannot be measured. For this reason, in order to evaluate thick steel material where the plastic deformation cannot be ignored, CTOD, which is a parameter applicable to material that can exhibit nonlinear behavior, was used in this study. The notch of the CTOD specimen was at the center of the weld metal and in the heat-affected zone. Specimens were prepared so that the notch was in the direction of the welding. According to British Standard (BS) 7448 Part 1: 1991, pre-cracking measured the fatigue crack length on one specimen as an actual value, which was derived from Equation (1). The actual crack length[mm] was calculated as the sum of the machining notch length [mm] and fatigue crack length[mm].

The fatigue crack length[mm] was measured at nine points to, located at a depth of 1% of the thickness from the surface and equally spaced between the two ends, as shown in Fig. 2. Excluding the end points, seven points from to were located at equal intervals inside the surface.

The formula for calculating the fracture test for 0.2% proof temperature was derived from BS 7448 Part 2 : 1997, as shown in Equation (2)^[8], where T is the test temperature, is the fatigue cracking temperature, and is the breaking test.

The equipment used a three-point bending tester DYHU-1000[kN] Universal machine to select the type of failure determined in Annex Fig. 13 of 1991 (characteristic types of force versus displacement records in fracture test). A temperature of -10±2[°C] was maintained for 38.5[min] as a condition for tempering. A mixture of LN, alcohol, and dry ice was used as the cooling medium.

3. Experiment result

3.1 Fatigue pre-cracking and fracture toughness of weld metal

In performing a general fracture toughness test, measurement of the pre-crack length was omitted because the surface crack length of the specimen was measured in the calibration step.

However, in this study, fatigue cracking occurred and grew in tempered steel with a thickness of 80[mm] or more, and pre-fatigue crack measurement was performed in consideration of the tunneling effect. Fig. 3 shows the details of the notch position of the welding joint.

Table 4 shows the results of the fatigue pre-cracking parameters at the weld metal area.

To measure the maximum fatigue stress intensity factor range and the stress ratio, applied loads were set to a force of 75[kN] in plastic deformation, using a stress ratio of 0.1, and a frequency of 10[Hz] in order to evaluate the effect of the frequency of the cyclic stress. The degree of influence related to whether or not cracks are formed was 151[MPa]. Table 5 shows the difference between the maximum and minimum values of the nine crack lengths.

3.2 Fatigue pre-cracking and fracture toughness in the heat-affected zone

In general, for high-stiffness steels, changes in strength, ductility, and toughness in the heat-affected zone showed different patterns from those of ordinary maximum and minimum values of the nine crack lengths. In the heat-affected zone, the overall strength was increased, but the area heated near the transformation point tended to have a lower strength than the base material. It showed the lowest ductility in the bond area, then recovered gradually and showed a higher peak value than the ductility of the base metal near the transformation point. It also had the characteristic of recovering to normal afterwards. The changes in toughness showed in the form of aging and embrittlement near the bond, where the grains were coarsened at transformation points. In addition, owing to the characteristics of the thick plate used in this study, the 2D flows were affected by the thickness and 3D heat flow state, where the plate thickness could be regarded as infinite. In 2D heat flow, defects such as torsion, deformation, and melting can occur because overheating occurs during welding. For this reason, pre-heating was performed before welding to slow down the cooling rate and prevent unnecessary embrittlement in the weld metal and heat-affected areas. The results of changing each variable of the fatigue pre-cracks in the heat-affected zone of the thick reinforced steel are also shown in Table 4.

The experiment was conducted under the same conditions as those used for the weld metal. The heat-affected zone also showed a similar pattern to the weld metal part, but the maximum stress factor, which is the degree of influence on the formation of cracks, was 156[MPa], which is slightly higher. Table 6 shows the maximum and minimum fatigue crack values obtained by measuring the crack length at nine equally spaced intervals.

3.3 Fracture toughness of weld metal and heat- affected zone

Fig. 4 shows the CTOD weld metal results. When the permanent displacement value of the notch is 3.27 [mm] and the maximum force is 291.5[kN], the CTOD is more than 1[mm].

There was no pop-in phenomenon, in which materials passed through the surrender store and showed a sudden drop in load and an increase in displacement. In this case, the mean values are prior to destruction at the local point of the initial crack shear. The change in crack length makes a difference in the binding force of the crack tip; therefore, even in the vicinity of the same weld, different results can be obtained depending. The CTOD results for the heat-affected zone are shown in Fig. 5. The curve indicated by the low-power regression line on the graph is the crack-resistance curve. This curve shows the area where the cracks are stable and no instability occurs. If the material is large, it can be seen by the slope of the curve that the resistance to propagation of the crack is high. Therefore, according to the slope curve, it is found to be an important characteristic value of the material.

During the test, the graph showed no unstable fracture, and the crack progressed relatively stably. Defining the maximum load at this time, the COTD value of the weld metal was 1.001[mm], and the CTOD value at the heat-affected zone was 1.527[mm]. Both welds largely satisfy the minimum CTOD requirement of 0.15[mm] of DNV and GL Classification Rules. However, it can be seen that the weld metal portion was significantly lower than the heat-affected portion. This is because the shape of the beveling at the welding joint and the force of shrinkage due to the binding or cooling of these shapes was affected. In particular, this experiment was conducted with the possibility of presence of a material that could cause defects, due to the accumulation of inclusions of microscopic foreign matter in the molten paper and slag generated during welding at the crack tip of each welding area. The final results of the fracture toughness in the weld metal and heat-affected zone are compared in Table 7.

4. Conclusion

In the case of thick steel plates and extremely thick steel plates, even welds that have passed non-destructive tests could still have very fine weld defects. Based on findings that show that these defects adversely affect the structure when a load is applied for a certain period of time, this study examined how high heat input affected each welding site at the welded joints of high-strength thick steel plate, which is a type of structure that constitutes the upper deck of a large container ship through fracture toughness tests, the following conclusions were obtained.