Abstract
To investigate the effectiveness of a viscous fluid damper (VFD) in mitigating fatigue damage in welded beam-to-column connections, six specimens were manufactured with the full penetration groove weld process. These specimens were divided into three groups: S1, S2, and S3, each containing two specimens. One specimen in each group was equipped with VFD, while the other was not. All specimens underwent the same elastic cyclic loading stage, but a different plastic cyclic loading stage for each group. The study results indicate that the presence of VFD can significantly enhance the load-bearing capacity of the welded beam-to-column connection by 10%–15% when subjected to the same displacement amplitude at the beam end. The primary cause of failure in the welded beam-to-column connection is the fatigue damage of the weld toe on the beam flange, which requires more attention during the welding process. Additionally, the VFD can effectively reduce stress concentration at the weld seam area, leading to a maximum of 50% decrease in the maximum stress near the weld toe of the beam flange center, as observed in our study.
Keywords
Introduction
Welded beam-to-column connections in high-rise steel buildings are susceptible to fatigue damage under cyclic loading in the form of high-cycle fatigue damage under wind and low-cycle fatigue damage under strong earthquakes. The fatigue damage may even lead to severe collapse of the whole building. Under such circumstances, it is indispensable to decrease fatigue damage under wind and earthquakes during the design of the welded connections. Traditionally, engineers are prone to increase the strength of welded connections to mitigate fatigue damage, but at expensive costs. Therefore, using vibration control technology to decrease fatigue damage in high-rise buildings has received more attention.
Vibration control technology is usually used to decrease the vibration effect of buildings. However, some researchers have started to notice that vibration control technology can be used to reduce fatigue damage. The research started mainly in the recent two decades. Palmeri and Ricciardelli (2006) estimated the fatigue life of structural components of tall buildings provided with viscoelastic dampers. They found that viscoelastic dampers can effectively mitigate the buffeting response and fatigue damage in medium-rise and high-rise buildings. Golafshani and Gholizad (2009) evaluated the efficiency of optimally designed tuned mass damper for fatigue damage mitigation in steel jacket platforms. They found that TMDs have shown an excellent performance in this application. Saiprasertkit (2014) conducted a comprehensive study on the fatigue strength assessment of load-carrying cruciform welded joints, focusing on the low- and high-cycle fatigue regions. The study introduces a new concept of effective notch strain and examines its applicability in both elastic and elastoplastic conditions for these joints. Ambrosio et al. (2014) proposed active control designed to minimize fatigue damage on the structure from the theoretical, numerical, and experimental points of view, and it showed a significant improvement. Ripamonti et al. (2015) proposed an adaptive vibration controller to increase the fatigue life of a smart structure made of composite material and actuated with piezoelectric patches. Debnath et al. (2016) studied the applications of tuned mass damper (TMD) systems for bridge structures in the mitigation of problems related to excessive vibration induced by either wind loading or vehicle loading, where dominant modes are usually in the vertical direction. Russell et al. (2020) described the development of a new type of steel beam with a unique dog-bone-shaped cross-section, named the Dinobeam. And conducted experimental and numerical investigations of the new steel beam subject to concentrated flange loading. Ghaedi et al. (2021) presented an experimental investigation of a novel yielding damper device, known as the Bar Damper (BD), designed for vibration control in structures during seismic events. The BD is composed of solid bars sandwiched between two plates and is intended to provide stable hysteretic behavior under cyclic loads, with the capability of undergoing large displacements without significant strength or stiffness degradation. Zare Golmoghany and Zahrai (2021) proposed a novel hybrid control system designed to enhance the seismic performance of structures. The proposed system, named Hybrid Friction Damper - Vertical Shear Panel (HFD-VSP), is a series combination of a friction damper (FD) and a vertical shear panel (VSP). It is designed to activate the FD during moderate earthquakes and both the FD and VSP during severe earthquakes, providing a two-level energy dissipation mechanism. Chandramohan et al. (2022) provided a comprehensive review of research on the static strength, fatigue life, and thermal performance of fillet welded joints, which are commonly used in various engineering applications such as buildings, bridges, railways, ships, and marine structures. Taheri et al. (2023) discussed the use of fillet and partial (incomplete) penetration butt welds in structural steel connections for seismic-resistant systems. It highlights that appropriately sized and executed double-sided, balanced fillet welds and partial penetration butt welds can offer performance equivalent to complete penetration butt welds under both semi-static and low cycle fatigue (seismic) loadings.
However, there are still limitations in the above literature survey. Firstly, the above literature discussed the fatigue damage mitigation of several kinds of structures, ranging from tall buildings, jacket platforms, and railway bridges; none of these researches discussed the fatigue damage mitigation of more complex welded components and structures. Secondly, most of the above research dealt with the problems from the theoretical or numerical point of view and lacked experimental verification. Finally, the above research only focused on one or two kinds of vibration control technology and lacked a systematic comparison between most kinds of common vibration control technology and its technical parameters.
So far, the authors have conducted some research in the field of fatigue assessment of welded beam-to-column connections before this research. The novelty of this research is to reveal the effectiveness of the VFD on fatigue damage mitigation in welded beam-to-column connections based on three comparative experiments of six welded beam-to-column connections with or without a VFD.
Experimental study
Overview of the fatigue test
The Material Properties of Structural Steel Q345.
Outline dimensional drawing (Fang et al., 2022).
Viscous fluid damper diagram. (a) Outline dimensional drawing and (b) Real product.
Specimen installation and loading diagram.
As shown in Figure 4, Six specimens of beam-to-column connections were manufactured with identical dimensions, loading positions, and constraint patterns. These specimens were divided into three groups: S1, S2, and S3, each containing two specimens. One specimen in each group was equipped with VFD, while the other was not, as listed in Table 2. The loading process comprised two stages: elastic cyclic loading and plastic cyclic loading. All specimens underwent the same elastic cyclic loading stage, controlled by forces with amplitudes of 60 kN, 100 kN, and 140 kN. Due to time and loading limitations, each amplitude of cyclic loads lasted only nine cycles with three specific loading rates: 36 s/cycle, 24 s/cycle, and 12 s/cycle, and each rate lasted three cycles, as listed in Table 3 and Figure 5. The plastic cyclic loading stage was controlled by displacements with amplitudes of 14 mm, 21 mm, and 28 mm and stress ratios of −1, 0, and −1 for groups S1, S2, and S3, respectively. The plastic cyclic loading stage was continued until the specimens ultimately failed or a maximum of 462 cycles with a constant loading rate of 24 s/cycle, as listed in Table 4 and Figure 6. The data, including the force of the MTS actuator, the displacement of the loading point, and the stress distribution in the weld area of the beam flange, were collected at a particular sampling frequency of 10 Hz, 10 Hz, and 20 Hz, respectively. The primary goal of the experiment on alleviating fatigue damage in welded beam-to-column connections is to make the specimen subject to as many cyclic loadings as possible before fatigue cracks near the weld toe are observed. The six welded beam-to-column specimens. The Details of Specimen Grouping. Elastic Cyclic Loading Pattern. Elastic cyclic loading pattern. Plastic Cyclic Loading Pattern. Plastic cyclic loading pattern.
The details about the experimental setup employed to examine fatigue damage and the efficacy of passive vibration control are as follows: First, we painted the welded area with white paint to make the crack be detected more easily, then, during the test, we shot the video on the welded area with a camera. The time that the crack started to initiate can be detected by the video, which was used to flag the force-displacement data recorded by the MTS actuator. Then, the fatigue life of the specimen can be calculated which is used to check the efficacy of the passive vibration control device.
Fatigue damage analysis of the plastic stage
Failure description
Specimen S1-1
The local fatigue crack was found on this specimen at the 58th cycle of the plastic cyclic loading stage, where the crack appeared near the weld toe on the lateral side of the beam flange. After a few cycles, the fatigue crack started to appear near the weld toe at the center of the beam flange, as shown in Figure 7(a). Meanwhile, another crack was found at the weld seam near the welding hole of the beam web. As the plastic cyclic loading continued, the two cracks at the lateral side and the center of the beam flange started to grow along the weld seam and finally merged into one crack, as shown in Figure 7(b). Meanwhile, the crack at the weld seam near the welding hole began to grow towards the remote side, as shown in Figure 7(c). The existence of the fatigue crack decreased the effective cross-section of the connection, which led to the rapid growth of the crack. At the 99th cycle, a fatigue crack was found on the other side of the beam flange. Finally, all the cross-section of the weld connecting the beam flange and the column was damaged, and the specimen was considered in a failure stage. Fatigue crack on specimens of group S1. (a) The cracks initiated near the weld toe on the beam flange of specimen S1-1, (b) The crack grew along the weld seam on the beam flange of specimen S1-1, (c) The crack initiated near the weld toe on the beam web of specimen S1-1 and (d) The crack initiated on the weld seam of the beam flange of specimen S1-2.
Specimen S1-2
The loading pattern of Specimen S1-2 with VFD was the same as Specimen S1-1. A slight fatigue crack was finally found on the weld toe at the center of the whole weld seam before the completion of the loading process. The crack grew at a very low speed and did not penetrate the width direction all along, as shown in Figure 7(d).
Specimen S2-1
The specimens in group S2 were subjected to unidirectional cyclic loads with zero stress ratio during the plastic cyclic loading stage, resulting in cracks appearing only on the tension side. The fatigue crack was first initiated on the weld toe at the lateral side of the beam flange weld seam at the 18th cycle of the plastic cyclic loading stage, and it gradually grew along the weld seam, as shown in Figure 8(a). The crack on the web plate soon followed, and it grew at a quite high speed, as shown in Figure 8(b). The test stopped when the cracks completely penetrated the cross-section area. Fatigue crack on specimens of group S2. (a) The crack initiated on the weld toe at the lateral side of the beam flange of specimen S2-1, (b) The crack quickly grew along the beam flange weld seam of specimen S2-1 and (c) The slight crack initiated on the beam flange weld seam of specimen S2-2.
Specimen S2-2
The same phenomenon that the crack only appeared on the tension side was found on Specimen S2-2. Its bearing capacity was remarkably increased due to the existence of VFD. The fatigue crack initiated near the weld toe on the beam flange and did not penetrate the whole cross-section even at the 462nd cycle after the test continued for 2 hours, as shown in Figure 8(c). Since the test focused on the fatigue initiation life of the specimen, the test finished after the crack was found.
Specimen S3-1
The fatigue crack was first found at the intersection location of the welding hole and the beam flange at the 7th cycle of the plastic cyclic loading stage, and it soon penetrated the flange thickness and grew towards the two lateral directions, as shown in Figure 9(a). Meanwhile, another crack was found at the lateral surface of the weld toe on the beam flange. It continued growing along the weld seam until it was 30 mm away from the lateral surface of the beam flange, then it changed the growing direction upward along a 45° angle and intersected with the crack first found. The crack on the beam web plate appeared at the 10th cycle of the plastic cyclic loading stage and grew at a tremendous speed, as shown in Figure 9(b). The final failure was caused by the damage of the beam web plate. Fatigue crack on specimens of group S3. (a) The crack initiated and grew on the beam flange of specimen S3-1, (b) The crack grew along the weld toe on the beam web of specimen S3-1, (c) The crack initiated at the weld seam near the welding hole of specimen S3-2, (d) The crack initiated at the opposite side welding hole of specimen S3-2 and (e) The crack grew through the weld toe on the beam web of specimen S3-2.
Specimen S3-2
A small fatigue crack was first found at the weld seam near the welding hole at the 9th cycle of the plastic cyclic loading stage. Only one cycle later, another crack was found on the other side of this weld seam, as shown in Figure 9(c) and (d); meanwhile, the weld toe on the beam flange suffered fatigue cracking. The crack growth rate was slower than that of S3-1, and the bearing capacity decrease was relatively steady. The final failure was caused by the fracture of the weld seams on the beam flange and beam web, as shown in Figure 9(e).
From the above descriptions, it is summarized that the specimens resisted the cyclic loading in the way of plastic deformation, and the final failure mode was the fracture of the weld seams. It is worth noting that the damage locations were mainly found at the weld toes on the side of the steel beams, and the steel beam flanges under tension stress were more susceptible to fatigue damage. Moreover, most of the fatigue cracks were first found at two locations near the weld seam on the beam flange: the lateral surface of the weld seam on the beam flange and the central part near the welding hole. In contrast, the fatigue crack found on the beam web often originated from the welding hole.
Based on the comparison between specimens with or without VFD, it is obvious that VFD can improve the stress concentration near the weld. For example, it was found that the complete cross-section fracture was not found on specimens with VFD in the previous two groups, and the fatigue crack growth rate was slower than that on specimens without VFD. This phenomenon can be explained by two reasons. The first reason may be that the damping force provided by VFD can decrease the stress amplitude near the welding location. The second reason may be that VFD restrains the buckling and plastic deformation of the beam flange.
Force-displacement hysteresis curve
The force-displacement hysteresis curve reflects the dissipated energy caused by the plastic strain under plastic cyclic loading, where a larger envelope area surrounded by the curve indicates better energy dissipation capacity. The hysteresis curves of all six specimens are illustrated in Figure 10, where the ordinate is the output force by the MTS actuator F, and the abscissa is the exerted displacement amplitude d. Force-displacement hysteresis curve for all specimens. (a) Specimen S1-1, (b) Specimen S1-2, (c) Specimen S2-1, (d) Specimen S2-1, (e) Specimen S3-1 and (f) Specimen S3-1.
From Figure 10, it is found that VFD significantly increases the ultimate bearing capacity of the specimen. The decrease of the bearing capacity for the specimens with VFD is slower than that of specimens without VFD. Moreover, compared with specimens without VFD, the maximum actuating force of specimens with VFD increases by 10%∼15% under the same displacement amplitude loading, which leads to a larger envelope area and better energy dissipation capacity. Therefore, more significant energy input is required for specimens with VFD to suffer fatigue damage.
The area of the hysteresis loops is relatively small when subjected to cyclic loading with a displacement amplitude of 14 mm and 21 mm as compared to that of 28 mm. This is mainly because, for the first two cases, the plastic deformation is quite small, and the energy is primarily dissipated by the plastic deformation of the weld seam at the connection area; however, for the third case, the plastic deformation is large enough, and the energy is primarily dissipated by the plastic deformation of the beam material.
It is observed from Figure 10(f) that the hysteresis loops decrease after a certain number of loading cycles. This could be due to the specimen being damaged during these loadings. The remaining energy dissipation capacity was then mainly provided by the VFD.
Stress analysis of the elastic stage
Stress Amplitude at the Center of the Weld toe on the Beam Flange.
The stress amplitude of the specimens without VFD is calculated by averaging the stress amplitude of nine cycles in each force amplitude, and the calculation for specimens without VFD is averaged by three stress amplitudes for each loading rate. The reason for processing data like this is that the VFD is sensitive to the velocity difference between its two ends, so different loading rates will lead to different stress reductions. It can be found that with the increase of the loading rate, the stress amplitudes of the specimens with VFD decrease in each force amplitude. This phenomenon indicates that the VFD is much more suitable to mitigate high-frequency vibration.
Compared with the specimens without VFD, the stress amplitudes decrease obviously for the specimens with VFD. This is mainly because the VFD can provide bracing force to the beam, reducing the force on the weld connection. This is beneficial for decreasing the accumulation of fatigue and increasing the fatigue life of the weld connection.
Conclusion
The study experimentally investigates the ability of VFD to reduce fatigue damage using six elaborate welded beam-to-column connections. Comparing the experimental phenomena and data between the specimens with VFD and without VFD, some meaningful conclusions can be made as follows: (1) The existence of VFD can significantly increase the load-bearing capacity of the welded beam-to-column connection by about 10∼15% under the same displacement amplitude acting on the beam end. (2) The failure of the welded beam-to-column connection is mainly caused by the fatigue damage of the weld toe on the beam flange which should be paid more attention to during the welding process. (3) The stress concentration at the weld seam area can be remarkably alleviated by the VFD with max to 50% decrease of the maximum stress near the weld toe of the beam flange center in our study. As the decrease of the stress amplitude can reduce the damage accumulation, the VFD is a good choice to extend the fatigue life of the welded beam-to-column connection.
It should be noted that this research only focuses on the effect of VFD on the fatigue mitigation of an isolated welded beam-to-column connection, it may have no significant effect on the fatigue mitigation of the whole high-rise steel building, which will be studied in detail in our future research.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Natural Science Foundation of China (52008202).
