استفاده از تجزیه و تحلیل حساسیت گشتاور برای دسترسی به شرایط جوشکاری / پردازش اصطکاک

The use of Friction Stir Processing (FSP) techniques for the joining and/or transforming of metallic materials is being object of intensive research since the earliest development of the Friction Stir Welding (FSW) technology in 1991. Despite of this, an accurate understanding of the main welding/processing mechanisms and its relation with the process parameters is still missing. Current paper intends to provide some further insight on this subject by discussing the relations between processing parameters, classified as independent variables, and the corresponding welding results, classified as dependent variables, using torque sensitivity analysis. The relation between base materials properties, plate thickness, welding conditions and torque evolution were also explored, which constitutes a novelty relative to the previous studies on this subject.

Since its development in the nineties, the friction stir processing (FSP) techniques, such as Friction Stir Welding (FSW) and Friction Stir Surfacing (FSS), which were developed for joining and/or improving locally material properties by using solid state processing principles, are being developed and applied based on trial and error analysis for optimizing processing conditions. Actually, full understanding of FSW/P mechanisms and/or thermomechanical principles was not achieved yet, still missing important data relating process parameters and material properties with processing conditions and processed components mechanical and metallurgical characteristics. According to Colligan and Mishra (2008), such understanding should be helpful in deciding how to change process conditions to achieve desired effects, such as improving materials strength, eliminating weld/processing defects and transferring welding/processing procedures to new processing conditions and/or materials. Finding a process output parameter, enabling full control of processing conditions and insuring suitable welding/processing results, is also an important step in consolidating the widespread application of FSW/P techniques at the industrial level. Recently, Longhurst et al. (2010) proposed the use of the spindle torque registered by the welding machines as a process control parameter for FSW, in alternative to the commonly used plunge depth or vertical force controlled setups. According to them, using torque control, it is possible to adapt easily the weld process to changing workpiece characteristics, since it provides a more suitable indicator of the tool depth into the workpiece than axial force. Pew et al. (2007) argue that registering the torque during welding, and using it for evaluating the heat input during the process, not only enables to avoid the difficult and time consuming task of setting thermocouples before welding, but also enables post-weld analysis of the temperature fields. Khandkar et al. (2003), for example, developed an extensive numerical study of temperature distributions and thermal histories during friction stir welding by using an input torque based thermal model. Some studies on the evolution of the spindle torque, with different processing variables, are also available in the scientific databases. Meanwhile, Cui et al. (2010) observed that torque depends mainly on rotational and welding speeds, Peel et al. (2006) and Arora et al. (2009) reported that the torque is relatively insensitive to the welding speed, since this parameter does not affect the temperature field as much as the rotation rate. Actually, several other works, such as that reported recently by Jacquin et al. (2011), clearly show that the spindle torque continuously decreases with increasing tool rotation speed. In all these works, the decrease in torque with increasing rotational speed was related to the decreasing flow strength of the base materials at increasing temperatures. This conclusion also points for a strong relation between torque and base material characteristics, which still needs to be explored. Colligan and Mishra (2008) showed that the welding torque also depends on tool parameters, such as shoulder and pin dimensions. However, the range of welding conditions tested was so limited that this aspect still needs further research. Actually, in spite of being well established the strong relation between torque and welding/processing conditions, all the studies reported to date are focused on a limited range of processing parameters, and most of them analyse results for a very specific material, tool geometry or plate thickness. Current paper intend to provide some further insight on this subject by discussing the relations between a broad range of processing conditions, classified as independent variables, and the corresponding welding results, classified as dependent variables, using torque sensitivity analysis. The welding conditions tested included not only testing varied processing parameters, which enabled to compare present data with previous studies from other authors, but also analysing the relation between base materials properties, plate thickness, welding results and torque evolution, which constitutes a novelty relative to the previous studies on this subject.

From present results it was possible to depict a strong influence of base material plastic behaviour and plate's thickness on welding results and on torque registered during welding. Despite of this, for constant tool parameters and axial load, the evolution of torque with tool traverse and rotation speed was found to be independent from both factors, being described by a unique power law equation. Modelling of torque results enabled to conclude that meanwhile the tool rotation speed and plate's thickness are the main factors in determining torque values, due to its strong influence on heat generation and through thickness heat distribution and material flow, the welding speed had no strong influence on average torque levels. Base material plastic properties also have a strong, but indirect, impact on torque results. Actually, plastic properties determine the sensitivity of the base materials, and consequently welding conditions, to varying tool rotation and traverse speeds, determining by that way torque results. Torque results were also found to be very sensitive to varying axial load, mainly in thicker plates welding, due to the above-mentioned reasons. No influence of varying shoulder diameter on torque results was reported.