Thermal-fluid-structure coupling analysis of void defect in friction stir welding

Highlights

• A novel integrated thermal-fluid-structure coupling model of the friction stir welding process was proposed for simultaneous prediction of the weld formation and tool service life.

• A new non-uniform distribution model was proposed to describe the interaction at the tool-workpiece contact interface.

• The severe reduction of tool-workpiece contact interfacial frictional shear stress is the main reason for the formation of the void defect in friction stir welding.

• The difference between the maximum and the minimum tool-workpiece contact pressure could serve as a numerical criterion to predict void defect formation.

• The tool is apt to fracture at its root under an improper welding condition since severe stress concentration is located there.

Abstract

Understanding the void defect formation mechanism and simultaneous predicting the tool service life in friction stir welding are critical for optimizing the welding parameters. However, the void defect formation mechanism in friction stir welding is not yet elucidated. In this study, a novel integrated thermal-fluid-structure coupling model of the friction stir welding process was proposed for simultaneous prediction of the weld formation and tool service life. A new non-uniform distribution model of the tool-workpiece contact pressure was proposed to describe the interaction between the tool and the workpiece. The void defect formation mechanism was quantitatively studied using the proposed integrated thermal-fluid-structure coupling model. The results show that the plastic material flows in the horizontal direction and can completely fill the cavity behind the tool for the welding condition of forming a sound weld. While the tool-workpiece contact interfacial frictional shear stress in the rear of the tool is decreased significantly which leads to a severe decrease in the plastic material flow velocity. Therefore, after bypassing the tool from the retreating side, the plastic material at the bottom of the weld stagnates, and void defect forms in the middle and lower part of the weld at the advancing side. The difference between the maximum and the minimum tool-workpiece contact pressure could serve as a numerical criterion to predict void defects. A sound joint is formed when the difference is lower than the critical value of 15 MPa, while a void defect is formed in the weld if it is higher than this critical value. The maximum equivalent stress acting on the tool is located at the pin root with severe stress concentration at a high welding speed. The front of the tool is subjected to tensile stress while its rear is subjected to compressive stress, therefore the tool is apt to fracture at its root under an inappropriate welding condition. The average normal stress of the tool varies periodically with its period consistent with the rotation period of the tool. The service life of the tool is decreased with the increase in welding speed and the decrease in rotation speed. The model is validated by experimental results.

Introduction

Friction stir welding (FSW), as a solid-state joining process, is capable to achieve a joint with excellent performance and is widely used for joining lightweight alloys such as aluminum and magnesium alloys in aerospace vehicles, automobiles, and railway trains [1], [2], [3]. During the FSW process, a non-consumable rotating tool is plunged into the workpieces to be welded. The interaction between the tool and the workpiece contact interfaces will generate frictional heat to soften the plastic material, as well as induce it to flow which result in severe plastic deformation heat [4], [5], [6], [7]. Thus, the interaction between the tool-workpiece contact interfaces determines the heat generation and plastic material flow behaviours in the FSW process, which finally determine the joint formation and mechanical properties [8], [9], [10], [11]. The welding parameters are crucial for the interaction between the tool-workpiece contact interfaces. Many studies have proved that employing improper welding parameters in FSW can lead to the formation of void defects in the joint which seriously deteriorate the mechanical properties of the joint [12], [13], [14], [15]. In addition, premature failure of the FSW tool may also occur under improper welding conditions [16,17]. Thus, it is of great significance to study the formation mechanism of the void defect in the FSW joint and reliably predict the FSW tool life for optimizing the welding parameters and controlling the joint formation.

Generally, it is difficult to directly observe and measure the internal void defects during the welding process. Therefore, non-destructive testing of the joints after the welding process is the most commonly used method to identify weld formation. For example, X-ray tomography [18,19] and ultrasonic testing [20,21] are employed to identify internal void defects in FSW joints. However, non-destructive testing of the joints after the welding process is time-consuming which seriously reduces the welding production efficiency. Furthermore, some researchers have explored detecting void defects through real-time monitoring of the welding temperature and force [22], [23], [24]. Das et al. [23] have proposed a method to detect the void defects in the FSW process using the measured welding temperature, it revealed that the void defects affect the temperature diffusion which changes the rate of temperature variation. Guan et al. [24] identified the internal void defect based on the feature of in-situ force signals, it revealed that the normalization ratio of the amplitude of the third harmonic and first harmonic of welding force can identify the void defect with an accuracy of 96.43%. Although experimental methods have demonstrated the possibility of identifying the void defects in FSW joints, the demand for high-precision measuring equipment also limits their application in industry. Therefore, Du et al. [25] investigated the conditions for void formation using a decision tree and a Bayesian neural network. They considered the unprocessed welding parameters and computed variables by FSW models as input data to predict the void formation, and it revealed that both the temperature and the maximum shear stress are the most important factors for void formation. Recently, Guan et al. [26] proposed a machine learning model to predict the void defects in FSW based on the measured force data which shows an accuracy of 95.8%. They found that the buildup of redundant material transported to the retreating side is the main reason for the characteristic variation of force when a void defect was formed. However, the currently available experimental data and machine learning models are still inadequate to fully understand the formation mechanism of void defects.

Numerical simulation of the FSW process is promising in elucidating the defect formation mechanism [27], [28], [29]. Schmidt et al. [30] developed a fully coupled thermomechanical three-dimensional model using the arbitrary Lagrangian-Eulerian (ALE) formulation in ABAQUS. The model allows for the possibility of a separation between the workpiece and the tool during the FSW process. Das et al. [27] developed a finite element model using a CEL approach to investigate the effect of process parameters on weld defects. The flow of material within and out of the solution domain allows the prediction of both the surface and subsurface defects by the FEM model. Except for the FEM model, the CFD model employs Eulerian mesh which is suitable for capturing the large plastic material flow in FSW. Zhu et al. [31] proposed a method to calculate the non-uniformly frictional force on the tool-workpiece contact interfaces based on iterative algorithms and the experimentally measured welding forces. It revealed that the non-uniform frictional force model is capable of predicting the void defects in FSW joints. Liechty et al. [32] proposed a variable shear stress boundary of tool-workpiece contact interfaces and revealed that the void defect is attributed to an insufficient tool-workpiece contact which leads to a vanishing of frictional force between the tool and workpiece. Ghat et al. [29] developed a two-dimensional numerical methodology to capture the material flow and predict the joint formation mechanism. It revealed that insufficient material filling leads to the formation of tunnel defects at a higher travel speed. Recently, Zhao et al. [33] proposed a pressure-dependent velocity boundary condition based on wear theory for modeling the self-reacting friction stir welding using the computational fluid dynamic approach and found the model is capable of capturing the weld formation characteristic. It also revealed that the tendency of the thermo-mechanical affected zone boundary moves toward the pin periphery at the advancing side could serve as a numerical criterion to predict void defect formation. Although some models have been proposed to predict the void defect in FSW. The critical thermo-mechanical state of plastic material flow during the formation of void defect has not been quantitatively analyzed. The effect of welding parameters on the tool-workpiece interaction and the formation mechanism of void defects is still unrevealed. Furthermore, there is still no characteristic parameter related to the welding parameters which could be used to conveniently predict the void defect in FSW.

Furthermore, the fracture of the tool due to severe welding forces is always accompanied by the formation of void defects at an improper welding condition [34,35]. The fracture of the tool not only reduces the service life of the tool and the welding efficiency but also increases the manufacturing cost, as well as reduces the stability of the welding process [36]. Therefore, it is of great significance to analyze the forces acting on the tool and predict the fatigue life of the tool for optimizing the welding parameters in the FSW process. Generally, it is difficult to measure and analyze the force acting on the tool exactly, especially the distribution of the force at the tool-workpiece contact interfaces [37]. Numerical simulation of the tool force in the FSW process is promising in predicting the tool service life and elucidating the influence of welding parameters on the welding force. Jaffarullah et al. [38] proposed a steady-state finite element model for analyzing the stress and strain at the FSW tool. They found that the critical points of the FSW tool are located mainly on the edge between the shoulder and the pin, which further leads to the failure of the tool. Huang et al. [39] proposed a computational fluid dynamic model to predict the maximal effective stress acting on the tool and evaluate the tool fracture due to the lack of strength. It revealed that the tapered thread pin with triple facets is more liable to fracture at a higher welding speed. Arora et al. [34] calculated the traverse force and torque during the FSW using a three-dimensional heat transfer and viscoplastic material flow model and evaluated the maximum shear stress experienced by the tool pin. It found that the severe stress and high temperature contributed to low values safety factors and short tool life in FSW of steel. Du et al. [40] employed multiple machine learning algorithms and a mechanistic model to identify the causative variables responsible for tool failure and found that the maximum shear stress is the most important causative variable for tool failure. Although efforts have been made to understand the welding force in the FSW process, the influence of welding parameters on the distribution of the force acting on the tool and the tool service life is still lacking.

To date, the interaction between the tool and workpiece contact interfaces during the friction stir welding process has not yet been systematically elucidated in the open literature, especially regarding coupled analyzing the heat generation and transfer, plastic material flow and tool force. Consequently, to quantitive analyze the influence of welding parameters on the thermo-mechanical coupled process in FSW for simultaneous predicting the weld formation and tool service life, an integrated thermal-fluid-structure coupling model of the friction stir welding process was established. A tool-workpiece contact pressure model related to welding parameters was proposed to capture the interaction between the tool and the workpiece. The formation mechanism of the void defect in the FSW joints has been elucidated and the influence of welding parameters on tool service life has been systematically analyzed. The relationship between the welding parameters and weld formation in FSW was quantitatively studied, and a characteristic parameter for predicting the void defect was proposed. It lays a solid foundation for optimizing the welding parameters to ensure the weld formation and tool service life.