Complete Joint Penetration Research Articles

Application of Convolutional Neural Networks for Classifying Penetration Conditions in GMAW Processes Using STFT of Welding Data

For manufacturing components with thick plates, such as in the heavy equipment and shipbuilding industries, the gas metal arc welding (GMAW) process is applied. Among the components that apply the thick plate GMAW process, there are groove butt joints, which are fabricated through multi-pass welding. Various welding qualities are managed in multi-pass welding, and the root-pass weld is controlled to ensure complete joint penetration (CJP). Currently, the state of complete joint penetration during root-pass welding is managed visually, making it difficult to confirm the penetration condition in real time. Therefore, there is a need to predict the penetration condition in real time. In this study, we propose a convolutional neural network (CNN)-based prediction model that can classify penetration conditions using welding current and voltage data from the root pass of V-groove butt joints. The root gap of the joints was varied between 1.0 and 2.0 mm, and the wire feed rate was adjusted. During welding, the current and voltage were measured. The welding current and voltage are transformed into a short-time Fourier transform (STFT) representation depicting the arc and wire extension lengths. The transformed dynamic resistance STFT information serves as the input variable for the CNN model. Preprocessing steps, including thresholding, are applied to optimize the input variables. The CNN architecture comprises three convolutional layers and two pooling layers. The model classifies penetration conditions as partial joint penetration (PJP), CJP, and burn-through, achieving a high accuracy of 97.8%. The proposed method facilitates the non-destructive evaluation of the root-pass welding quality without expensive monitoring equipment, such as vision cameras. It is expected to be immediately applied to the thick plate welding process using readily available welding data.

Complete-joint-penetration vacuum laser beam welding of nickel-based superalloy 718Plus

Deep-penetration welding of nickel-based superalloys is an indispensable manufacturing technique for aircraft engines, aerospace engines, and nuclear containers. In this study, a newly developed vacuum laser beam welding method was used to weld a 9 mm thick superalloy 718Plus. A defect-free complete joint penetration weld with a high depth-to-width ratio was fabricated. The macro/microstructures, microhardness, tensile properties, and fatigue properties of weld joints under both as-welded and heat-treated conditions were systematically examined. The weld joint showed a microstructure of planar, cellular, columnar, and equiaxed crystals from the fusion line to the center line. The post-weld heat treatment had an insignificant effect on the grain structure but changed the precipitated phase in the weld. The newly formed γ′ and η phases were observed in the heat-treated sample. The mechanical properties showed microhardness, tensile strength, elongation, and fatigue strength of 488.6 HV, 1347.0 MPa, 13.7%, and 399 MPa, respectively, which were 104.6%, 93.2%, 77.8%, and 74.2% of those of the base metal, respectively. Both as-welded and heat-treated samples exhibited ductile ruptures. The excellent mechanical properties of the heat-treated joints were attributed to the formation of precipitated phases.

Seismic behavior of an improved drilled flange connection for I-beam to CFST column

The use of cold-formed steel tubes in the concrete-filled steel tubular (CFST) columns makes the design of the moment connections between beams and CFST columns much more challenging. This is because setting the internal diaphragms in the CFST column to transfer the bending moment of the beam end becomes impossible, and thus new joints are needed. A novel joint was proposed in a previous study by the authors, where two end plates were used for the force transmission between CFST columns and beam flanges. Moreover, to satisfy the “weak member strong joint” requirement, the I-beams were reduced by drilling a series of circular holes in their flanges. However, most of the tested specimens failed by the unexpected rupture of the complete joint penetration (CJP) groove welds between the beam flanges and end plates. In this paper, three specimens are redesigned and tested, where modifications are made to improve their behaviors under the seismic load. In this test, the local buckling of beam flanges at the drilled sections was observed in all specimens, and two specimens failed due to the fracture at the drilled sections, while CJP weld fracture and significant bending of beam flanges both occurred at the failure of the rest specimen. The hysteresis curves of all specimens are fusiform, indicating their good energy dissipations (the maximum equivalent viscous damping coefficients between 0.35 and 0.40). The failure rotation angles are about 2.44–3.01, and the ductility coefficients are between 2.26 and 2.66. According to EC3, all specimens can be classified into the rigid and full-strength joints. Subsequently, the parametric analyses using the validated finite element model are carried out, where the effects of main parameters are investigated on the seismic behavior of the presented joint. This study brings insights into the moment connections between the steel beam and CFST column with the outer cold-formed section.

Experimental and numerical study on weld strengths of built-up steel box columns

This study investigates the demands generated in the panel zone of welded steel built-up box columns with fillet and complete joint penetration (CJP) groove welds, which are mainly used as columns in steel moment-resisting frames. The motivation for this study stems from the a 360–16 specification for structural steel building, which requires the joining of flange and web plates of box columns with CJP groove welds within a zone extending 12 in. above the upper beam flange and 12 in. below the lower beam flange in special and intermediate steel moment frames. Laboratory experiments have been conducted to examine the strength of these welds under pure shear forces. For this purpose, two laboratory specimens, one with fillet welds and the other with CJP groove welds, have been designed, and the strength of these welds were equal to each other in pure shear force conditions. The fillet weld specimen has resisted up to 1670 kN in the pure shear test, and the CJP groove weld has resisted up to 1520 kN. The entire loading process of each specimen was characterized by employing the digital image correlation (DIC) technique, which enabled the extraction of strain values throughout. Numerical modeling is validated in Abaqus software using experimental results, with comprehensive analyses on force-displacement diagrams and rupture. The outcomes of the tests conducted have shown that the demands generated in the panel zone of the built-up steel box welded sections are satisfactorily tolerated by both fillet and CJP groove welds. Research significanceThe significance of this study can be elaborated in multiple aspects. Firstly, it introduces an innovative test setup meticulously designed to replicate the behavior of welded built-up box columns with corner joints to steel plates. This advanced experimental arrangement not only facilitates the current study, but also serves as a versatile tool for conducting forthcoming experiments, allowing for meaningful comparisons and further exploration of diverse phenomena within the field. Secondly, the deliberate imposition of uniform demands on both types of welds during the study enables a comprehensive examination of their behavioral characteristics. This comprehensive analysis contributes to a more profound comprehension of welds performance under identical conditions, thereby enhancing our knowledge within the field. Finally, the research outcome, which reveals that fillet welds exhibit almost identical load-bearing capacity in comparison to CJP groove welds, bears substantial practical significance for the construction industry. This finding has the potential to instigate more cost-effective and efficient welding practices, which are paramount for maintaining the structural integrity of steel structures.In summary, this study makes a noteworthy contribution to the scientific community by presenting an innovative test setup, deepening the knowledge of weld performance, and suggesting practical enhancements in welding methodologies. These collective contributions serve to elevate the field of structural engineering, promoting advancements in construction practices.

Cyclic behavior of dual-steel beam-to-column welded flange-bolted web connections

Cyclic loading tests on dual-steel beam-to-column welded flange-bolted web connections were conducted to quantify their moment resistance, plastic deformation and energy dissipation capacities. The test program consisted of five one-sided connection specimens, where Q355 grade beams and Q690 grade columns were used, to study the effects of beam-to-column welded flange connection details and panel zone shear strength on the connection seismic performance. Two types of welded connection details were considered; one was the traditional complete-joint-penetration (CJP) welded connections by use of backing bars, and the other type was those with the bottom backing bar further reinforced by a fillet weld. Three column web thicknesses were designed to arrive at strong, intermediate and weak panel zones, respectively. It was found that the CJP weld with the backing bar reinforced only under the bottom beam flange produced satisfactory performance and accommodated plastic rotations larger than 0.03 rad, while a maximum plastic rotation of 0.04 rad could be developed in the high-strength steel panel zone before fracture occurred. These results evidenced that those dual-steel connections could still sustain high seismic deformation demands.

Experimental study on seismic behavior of cover-plate connections between steel beams and high strength steel box columns

As one of the beam-to-column fully restrained connections that could shift the plastic hinges of the beam away from the face of the column, cover-plate reinforced connections are expected to be a practical choice for high strength steel frame seismic applications. Five cover-plate reinforced and one unreinforced welded flange-bolted web connection specimens combined with Q355 steel I-shaped beams and Q460 or Q690 steel box columns were designed according to Chinese codes and tested under cyclic loads at full scale. Failure modes, including local buckling of the beam, complete joint penetration (CJP) weld fracture, and column web fracture, were reported. The strength, stiffness, deformation, and energy dissipation features were analyzed. The four cover-plate connections met the AISC requirements for special moment frames (SMF) with maximum story drift angles greater than 0.06 rad and exhibited satisfactory seismic behavior. According to the test results, a limit on the ratio of the diaphragm plate to column plate thickness should be considered when the electroslag welding (ESW) process is used in high strength box columns. The strong panel zone designed with a panel zone resistance ratio (γpz) greater than 1.20 was recommended for box columns with steel strength no less than 690 MPa. Additionally, other design implications of the results were discussed.

Hysteretic behaviour of welded connections with highly inelastic panel zones

AbstractThe current practice in capacity‐designed steel moment resisting frames (MRFs) worldwide allows for limited shear yielding in the column web panel zone. As such, inelastic deformations concentrate near the beam ends, thereby leading to flexural strength deterioration due to local buckling often at modest lateral drift demands. Experiments on post‐Northridge welded connections suggest that prior to a story drift angle of 5% rad, panel zone kinking does not induce fracture for panel zone shear distortions of up to 10γy. Within such a context, this paper presents an alternative design of a welded moment connection with highly inelastic panel zones. The connection features simplified weld details including a customised beveled backing bar that can be kept in place after the completion of the complete joint penetration welds at the beam flange‐to‐column flange joint. Full scale experiments suggest that the hysteretic response of the connection remains stable up until a story drift angle of 8% rad. The specimen did not exhibit visible structural damage prior to story drift angles of 4% rad, which are characteristic to a maximum considered earthquake. Modest local buckling occurred near the beam ends only after a story drift angle of 5% rad, which is a typical drift threshold in current collapse assessment methodologies. This paper summarises the primary findings of the experimental program for one of the test specimens.

Real-Time Penetration Recognition Based on Voltage Signal in K-TIG Welding

This paper focuses on discerning the melting state of keyhole tungsten inert gas (K-TIG) welding, a promising welding technique still considered a novelty. In order to achieve this goal, a sensing system is established to collect voltage signals during the welding process of 304 stainless steel. Analysis of the collected signals reveals that the characteristic signal power spectrum between 36 kHz and 37 kHz is closely related to whether the welding has complete joint penetration. The experimental findings confirm that the spectral power density of the characteristic signal varies significantly with the alteration of welding quality. Moreover, it is observed that when the welding melting state remains unchanged, the change in welding parameters is not noteworthy. When adjusting welding parameters to maintain stable feature signals during the welding process, the corresponding penetration state of the welding does change. Therefore, it is possible to ensure the stability of welding quality by altering welding parameters in real time. These research results provide valuable insights into the real-time quality monitoring of the K-TIG welding process.

Seismic retrofitting of beam to double-I built-up column moment connections by lateral T-stiffeners: Experimental and numerical study

In the present paper, the retrofitting of welded flange plate moment connections joined to double-I built-up columns is discussed by presenting the design method and details of construction retrofitting. Some of the usual construction problems of this connection include the defective complete joint penetration groove weld at the connection of the flange plate and the column, the large out-of-plane deformations of the column cover plate at the level of the beam-to-column connection, and as a result, its inappropriate seismic performance and the connection's partially restrained behavior. Thus, an attempt is made to evaluate the advantages of strengthening this defective connection with the help of lateral T-stiffeners using analytical and experimental studies. For this purpose, the experimental specimen and the high-fidelity finite element model are considered for both retrofitted and defective connections by evaluating the proposed design method, cyclic behavior of connection, collapse potential of weld fasteners, and failure modes. The results indicate that by using the proposed retrofitting method, the connection is classified in the group of fully restrained connections, while also having the necessary strength and ductility to accept the conditions of the special moment-resisting frames according to AISC seismic provisions. By performing a sensitivity analysis on the T-stiffener, changing its dimensions and examining the damage indices, the accuracy of the proposed design method is proven. The examinations indicate that the aforementioned retrofitting method is a proper method for the introduced defective connection.

Seismic shear behavior of new high-strength reinforced concrete column and steel beam (New RCS) joints

New high-strength reinforced concrete column and steel beam (New RCS) joints were developed in this research. Two-way, through-beam New RCS joint details with beams concentrically or eccentrically connected to the joint were proposed. The beam in one direction was continuous through the joint (continuous beam). The beam in the other direction was connected to the continuous beam by complete joint penetration (CJP) groove welds. Five-spiral reinforcement was used for joint transverse reinforcement. A method to calculate the amount of five-spiral reinforcement was proposed considering the confinement effect from the face-bearing plates (FBPs). Doubler plates were used to strengthen the eccentric beam flanges with predrilled holes to allow the passage of column longitudinal reinforcement. A method was proposed for the design of the doubler plates. Large-scale New RCS joint specimens designed to fail in joint shear were tested. Test results showed that the proposed joint details and design methods effectively avoided failure in undesirable locations. All the specimens failed in joint shear but still reached high drift ratios. The eccentric joint specimens (IDEHS and ISEHS) exhibited peak joint shears higher than the concentric beam specimen (IHS). The 1994 ASCE and 2015 ASCE shear strength models produced conservative predictions of the peak joint shear and the joint shear at 0.5% joint shear deformation for all the specimens. The predictions by the 2015 ASCE method were more reasonable with less scatter than the 1994 ASCE method because the former recognizes the shear strength contribution from the outer joint panel even without the use of shear keys and allows a longer steel web in shear strength calculation.

Influence Mechanism of Metal Vapor in Plasma Arc Lap Welding

A three-dimensional coupled tungsten electrode–plasma arc–metal vapor– weld pool model was developed to investigate plasma arc and metal vapor characteristics, weld pool convection, and energy transfer in conduction plasma arc lap welding. The arc energy efficiency was also calculated. The numerical results showed that in conduction plasma arc lap welding, the constraint effects of the plasma arc by a small constricting nozzle, plasma gas, and electromagnetic force were strong, and no keyhole was formed inside the weld pool, so the heat flux on the weld pool surface was high as well as the weld pool temperature and mole fraction of Fe vapor above the weld pool surface. The high concentration of Fe vapor in the arc decreased the conduction energy from the plasma arc to the weld pool along with the arc energy efficiency. The calculated arc energy efficiency was only 50.2%. Without considering Fe vapor, the calculated weld pool had complete joint penetration. When considering Fe vapor, the calculated weld geometry agreed well with the experimental result.

Experimental and analytical investigation of drilled flange connections (DFCs) with radial drilling patterns

The capability of reduced beam sections (RBS) to improve the seismic behavior of rigid connections leads to their notable application in steel moment-resisting frames. However, they may encounter shortcomings such as local buckling and lateral-torsional buckling. To resolve such problems, the idea of utilizing perforated beam flanges instead of cut-out flanges as reduced beam sections was presented. These type of connections can reduce the potential for local and lateral-torsional buckling of conventional RBS connections and improve their seismic performance. Achieving such advantages requires special care to geometrical specifications and drilling patterns (DPs). Appropriate arrangement of holes reduces the stress concentration around the beam flange-to-column Complete Joint Penetration (CJP) groove weld and, consequently, prevents the groove weld and inter-hole area from tearing and fracturing in early cycles of loading. The present study investigated the hysteretic performance of the drilled flange connections (DFCs) of various drilling arrangements through an experimental program.It should be noted that the box-shaped columns were considered in this study to target the 3D moment resisting frames (MRFs) that all their axes participate in the lateral force-resisting system. The most appropriate drilling patterns were selected using remarkable numbers of numerical analyses. Four laboratory specimens with different drilling arrangements of top and bottom flanges were fabricated based on preliminary numerical analyses. The laboratory specimens were subjected to cyclic loading. To evaluate the specimens' seismic performance, moment-rotation hysteresis curves were compared, and the stress distribution at the connection CJP groove welds. The results of the experimental tests revealed that those DFCs with combined DP with notches and holes (DFC1), radial DP (DFC2), semi radial DP with notches (DFC3), and semi radial DP (DFC4) could increase the rotational capacity of the DFCs up to 50% of the required rotation of the special moment-resisting connections. Moreover, semi radial DP (DFC4) may increase the connection's rotational capacity up to 75% of the special moment-resisting connections. Also, Based on the analytical results, conventional RBS connections may increase the 0.04 radians, needed rotational capacity up to 37.5% for a pre-qualified moment-resisting connection. It was also shown that no lateral-torsional buckling and smaller degradation in stiffness and strength took place in plastic hinge zones at the relative deformation of 0.04 radians. The plastic strain at the center of the CJP groove weld line in the beam longitudinal direction was more significant than those corners of the groove weld line in the same direction. In all specimens, the minimum plastic strain of the CJP groove weld took place in central parts of the weld in the transverse direction. In DFC1, DFC2, DFC3, and DFC4 specimens, the contribution of panel zone rotation in the overall plastic rotation of the connection is negligible.

Effect of Heat Input on LMHMW Joint of Carbon Steel

Laser-MIG hybrid multi-layer welding (LMHMW) technology has been employed in paraxial configuration with laser leading for the welding of 20 mm thick Q235 carbon steel plates to exploit the hybridization effect that addresses the shortcomings of the individual process as well as to compliment their merits. The bilateral effects of arc augmented laser welding have resulted in complete joint penetration, process efficiency, stability and gap bridge ability. Samples welded under varying heat inputs in multiple passes have been analyzed for their microstructure evaluation using an optical microscope followed by tensile and Vickers hardness testing in various regions of the weld zones. This process was conducted to characterize the effect of heat input on the mechanical properties of the welded joints. The experimental results illustrate that different heat inputs have significant effects on the microstructure, heat affected zone width and mechanical properties of welded joints. The microhardness near the fusion line decreases dramatically due to the influence of the phase transformation process, and the highest microhardness value is obtained in the center of the weld seam. By using reasonable process parameters, the strength of the welded joint can obtain 458.5 MPa.

Brittle fracture estimation at ends of groove welded joints under cyclic loading

To prevent brittle fracture in structural components during earthquakes, their cumulative ductility (plastic deformation capacity) should be determined beforehand and appropriate countermeasures applied accordingly. In this paper, a method for estimating the cumulative ductility of welded joints with defects based on the σW–Nf relationship and Miner's rule is proposed for steel structure components to facilitate brittle fracture estimation. To verify its effectiveness, it was applied to the testing of welded joints with defects and compared to experimental results. The experimental tests were conducted under cyclic loading conditions to reproduce brittle fractures starting from the ends of complete joint penetration groove welded joints. The joint models used in this study were part of beam-to-column connections between beam flanges and diaphragms. Defects were artificially introduced into the welded joints before the experimental tests; brittle fractures occurred at these preinstalled defects in the welded joint specimens. The calculated cumulative ductility was obtained from the fracture toughness and Weibull stress of the joint model measured using finite element analyses. The calculated cumulative ductility values corresponded well with the experimental results, albeit with some explicable errors. The results indicate that the proposed method can estimate the cumulative ductility for structural components with defects under cyclic loading independent of their materials, notch types, and connection details. Thus, the findings could contribute to the assessment of defects in welded joints in steel structures subjected to earthquake loadings.

Closed-loop control of medium-thickness plate arc welding based on weld-face vision sensing

In one-side welding with full penetration, the use of backing strips helps support molten weld metals, but increases manufacturing cycle time and cost. The backing-free welding technology is under investigation for medium-thickness structural components in the aerospace industry. Its crucial challenge is ensuring complete joint penetration and maintaining a suitable and stable backside bead width. This paper presents an innovative method of controlling weld formation based on weld-face vision sensing in arc welding. A standard laser triangulation setup is used to obtain images of weld pool, and a new weld-face feature ‘depression degree of the weld pool’ is proposed. An image processing procedure is developed to extract the depression degree of the weld pool from images in real-time. The depression degree of the weld pool is found to correlate closely with the backside bead width, and is a better indicator of the backside bead width than the weld pool width, with a higher consistency and a smaller time lag. Employing the depression degree as a feedback signal, a backside bead width control method is developed and implemented. Welding experiments on 6-mm-thick aluminum alloy 2219 plates show that the presented method is remarkably effective in achieving full penetration and keeping backside bead width constant.

Qualification of Electroslag Welds made from HPS 485W (70W) and 345W (50W) Steels

Electroslag welding (ESW) is a highly efficient welding process for completing single-pass complete joint penetration welds that offers production gains over conventional welding processes. This gain in productivity requires very high heat inputs, often to the detriment of the base metal's heat-affected zone (HAZ) impact energy. U.S. bridge-welding codes recognize the ESW Narrow Gap (termed ESW-NG) process developed in the early 1980s to remedy the substandard weld and HAZ impact energies found in bridges across the United States fabricated with the original ESW process used roughly from 1967 to 1977. ESW-NG process is limited to certain base metals; particularly, ESW-NG with quenched and tempered (Q&T), and thermo-mechanically controlled processed steels are prohibited because they are not part of the ESW-NG process development. This study was specifically performed to evaluate whether the existing ESW-NG process could be qualified for use with grade HPS 485W (70W) Q&T bridge steel and hybrid weldments between grade HPS 485W (70W) Q&T and conventionally processed grade 345W (50W) bridge steels. None of the weldments met fracture critical impact requirements. The weldment of just Q&T steel and the hybrid weldment met nonfracture critical requirements. However, the weldment of just Q&T had tension and impact resistance results right on the margin, leaving little room for error.

Investigating the seismic behaviour of high-performance steel plate shear walls

Steel plate shear wall (SPSWs) have good ductility, high strength and stiffness, and considerable energy absorption capability. These advantages have been confirmed in experimental and numerical studies. However, despite these advantages, the need for huge columns surrounding the infill plates is the main limitation of SPSWs. Although the high-performance steel plate shear wall (HPSPSW) has solved this shortcoming, there are still some unknowns about this high-performance system. The HPSPSW system is expected to be more economical than other resisting systems due to ease of fabrication, the elimination of complete joint penetration field welds, the elimination of stress transfer from infill plates to main columns, the higher modification factor and no need for huge main columns. A comparison of analysis results of HPSPSW and SPSW systems revealed that the HPSPSW has desirable stiffness, ductility and energy absorption when its length to height ratio is greater than 1.5. In addition, the HPSPSW system had an ultimate lateral strength only 2–5% lower than that of the SPSW. Modification factors were calculated and average values of 6.67 and 8.64 were obtained for the SPSW and HPSPSW systems, respectively. Therefore, a modification factor of 8.5 is suggested for the HPSPSW.

Structural Performance Evaluation of CFT Columns with Built-up Square-shaped Steel Tubes using Production Methods

This study obtained compression test results and macro examination results for concrete-filled steel tube (CFT) columns with built-up square-shaped steel tubes to evaluate both the structural performance and the possibility of using the current CFT design for changes in production methods. The CFT columns have three variables, namely, welding details, welding types, and steel bar types, of which welding details are the main variable. The compression test results were compared with the nominal compressive strength, Pn, based on the material test results and complete joint penetration (CJP). The test results indicated that the ratio of the experimental maximum load to the theoretical calculation result was between 1.04 and 1.12 (1.08 on average). This means that it is possible to use the current CFT design in the current Building Structure Standards (KBC 2016, KDS 41 31 00). The macro examination results indicated that the quality of welding will be improved owing to the minimization of defects during welding if the groove angle is improved to 50°.

Effect of weld specifications on seismic performance of built-up box columns

Application of built-up box sections as columns of steel moment frames is very favorable due to their bi-directional flexural resistance. According to seismic provisions, for built up box columns, within a zone adjacent to the beam flange, flange and web plates of box columns shall be joined by complete joint penetration (CJP) groove welds. In spite of the fact that fillet welds are much easier to execute and are economically more efficient, they are prohibited within the panel zone of columns. In this paper the effect of weld specifications on seismic performance of panel zone in built-up box columns with corner welded joints is experimentally investigated, mainly focusing on the comparison of performance of fillet and groove welds used for fabrication of this type of columns. Based on the results of the experimental program on 1:2 scale specimens, representing beam to column connections of steel special and intermediate moment frames built using fillet and groove welds in panel zones, it is concluded that the performance of fillet welds could be as good as CJP groove welds in panel zones of built-up box columns for the qualifying drift angle of steel moment frames.

Numerical Study of Panel Zone in a Moment Connection without Continuity Plates

ABSTRACT Built-up box columns are more frequently used in earthquake-prone regions such as Japan and Iran. As the installation process of continuity plates and welding inspection are difficult, expensive, and time-consuming, this study was aimed at experimentally and numerically investigating continuity plate elimination from built-up box columns with regard to the seismic design specifications of special and intermediate moment resisting frames. Results of an experimental study of three full-scale rigid connections between I-beam and box-column subjected to cyclic loading were used for comparison and verification of numerical results. Experimental study showed that samples reached story drift angle of 0.06 radians before experiencing the permissible strength degradation. Therefore, the tested connections satisfied the criterion for special moment resisting frame (SMRF) in AISC. Based on observations of crack initiation and propagation in the experimental study, a rupture index (RI) was employed to investigate the effects of different complete joint penetration (CJP) groove welding methods used in the experimental study on connection ductility as well as to identify the regions prone to crack initiation and propagation. Both the numerical and experimental results showed the superiority of using back plate in CJP welding.