RS/AS AND NUGGET ZONE SEM ANALYSIS AND MATHEMATICAL EQUATIONS FOR PARAMETER OPTIMIZATION ON FRICTION STIR WELDING TOOL

Yugandhar Meka ¹; Prabhakar Kammar ²

ABSTRACT

In order to find the ideal process parameters, the friction stir welding (FSW) production process is combined with mathematical optimisation techniques in the current thesis. In terms of optimisation, the goals (or objec-tives) are process- related in that they define or represent how the process behaves. The Welding Institute (TWI),UK introduced the FSW process, which has been enhanced in this study, as a relatively new welding tech-nology in 1991. In conclusion, because the process is solid-state and does not involve melting, many of the problems of conventional fusing welding techniques may be avoided. In the FSW process, two work pieces are submerged in a spinning tool. As a result of friction and plastic dissipation, the temperature rises to the point where the material is sufficiently softened to be mixed together, forming a weld.Involving solid material ow, heat transmission, thermal softening, recrystallization, and the production of residual stresses, the process is mul-tiphysics in nature.

Keywrods: magnesium alloys, friction stir welding tensile strength, tensile elastic limit, Young’s modulus, Poisson’s ratio, yield strength, strain-hardening characteristics, AZ91.

Recibido: 03 de septiembre de 2023Aceptado: 20 de diciembre de 2023

Received: September 03, 2023Accepted: Accepted: December 20, 2023

Cómo citar este artículo: M.Yugandhar, Prabhakarkamma. RS/AS and nugget zone SEM analysis and mathematical equations for parameter optimization on friction stir welding tool”, Revista Politécnica, vol.20, no.39 pp.78-98, 2024. DOI:10.33571/rpolitec.v20n39a6

1. INTRODUCTION

---

The purpose of this work is to use numerical modeling and optimization tools to study and control thermo-mechanical conditions, ie residual stresses and workload conditions, in friction stir welding (FSW) process-es. Since the 1990s, various theoretical/experimental studies and EU-funded industry initiatives such as JOIN-DMC and DEEPWELD have been conducted to better understand the many physical components of the FSW process. rice field. FSW is already used in regular and critical applications for joining structural components, mainly made of aluminum and its alloys. However, further research is still being done to achieve a more robust "process window".

FSW modeling includes a wide range of models such as microstructural evolution, material flow, heat flow, heat generation, residual stress, mechanical loads (such as tool forces) and bond strength. To gain a deep-er understanding of some of the basic mechanisms of the FSW process, the physics is divided into sub-systems, namely thermal and mechanical behavior, and then integrated modeling approaches (1–5 ) is being performed. Compare these two physical aspects with the actual load cases. Microstructural evolution, mate-rial flow, heat flow, heat generation, residual stresses, mechanical loads (such as tool forces), and bond strength are all part of FSW modeling. In order to better understand some of the underlying mechanisms of the FSW process, we divided the physics into subsystems, namely thermal and mechanical behavior, and used an integrated modeling approach to study these two physical aspects. combined the combined be-havior of the with the actual behavior. load case.

Most of the work done in the current research study has been devoted to computer models of various components of the FSW process (6). In particular, heat diffusion and transient/residual stresses occurring during and after welding were investigated numerically and compared with published experimental data. Therefore, important information on the development of residual stresses as a function of key FSW pro-cess parameters such as tool rotation speed and cross-welding speed was collected using the general-purpose Nite Elements (FE) software ANSYS (Ansys Inc., 2007). will be, ABAQUS (Dassault Systèmes, 2007) and COMSOL (COMSOL AB, 2007) [8].

Following our modeling studies, we used numerical optimization techniques, including classical and evolu-tionary algorithms, to find the optimal process parameters.

2. EXPERIMENTAL WORK

---

FSW is a tool with a threaded/unthreaded cylindrical/probe shoulder [9]. As shown in Figure 1.1, the probe (pin) rotates at a constant speed and travels at a constant travel speed in the connecting line between the two butt- welded sheets or plates of material. The part must be firmly secured to the backing plate to sup-port the high plunge pressure of the FSW machine head (7) and prevent the adjacent mating surfaces from being pushed apart. The pin length should be slightly less than the desired weld depth and the tool shoulder should be in direct contact with the workpiece surface. Insert the probe into the workpiece and move the tool along the weld at a 2-4 degree bevel angle to increase pressure under the tool shoulder.

Fig .1. A schematic diagram of taper tool

Fig .2. A diagram of the FSW process

Frictional heat is generated between the wear- resistant welding tool shoulder [10] and the pin, and on the other hand with the work material. This heat, combined with the heat generated by plastic dissipation during the mix-ing process, softens the agitated components without reaching their melting point (hence the name solid-state process), leaving the tool in a plasticized tubular shape. Allows crossing metal weld lines. When the pin is pushed in the welding direction, the front of the pin pushes the plasticized material into the back of the pin, applying a large forging force to strengthen the weld metal [11].

Welding of materials is facilitated by strong plastic deformation in the solid state accompanied by dynamic re-crystallization of the weld nugget.

Figure.3.(a) Frictiction stir welding on magnesium sheets

Figure 3.(b) H13 tool steel tool pin and shoulder

The solid-state nature of the FSW process, along with its specific tooling and asymmetric nature, results in a very unique microstructure [12]. Certain areas are common to all weld types, while others are unique to her FSW process. Nomenclature varies, but the following reflects general consensus.

The stirring zone (also called nugget or dynamic recrystallization zone) D in Figure 1.3 is a highly deformed re-gion of material that roughly corresponds to the position of the pin during welding [13]. Particles in the stir zone are nearly equiaxed and are often ten times smaller than those of the starting material (Murr et al., 1997). The extensive presence of many concentric rings in the stirring zone is called the onion ring structure. The spe-cific origin of these rings has not yet been determined, but differences in particle number density, grain size, and organization have been suggested [14].

The flow arm is at the top of the weld and consists of material that is pulled by a shoulder from the retracting side of the weld behind the tool and deposited on the advancing side of the weld [15].

The thermo-mechanically affected zone (TMAZ), represented by the letter C in the figure. 1.3, located on both sides of the stirring zone. The stresses and temperatures in this region are lower than in the stir zone, so the microstructural impact of welding is also less. In contrast to the stir zone, the microstructure is clearly recog-nizable as that of the starting material, albeit clearly distorted and rotated. Although the term TMAZ theoretically refers to the entire deformation area, it is commonly used for areas not yet covered by the terms 'touch zone' and 'forearm' [16].

A heat affected zone (HAZ), represented by B in Figure 1.3, exists in all welding processes.

As the name suggests, this zone is subject to thermal cycling but does not deform during welding. Although the temperature is lower than TMAZ, it can be affected if the microstructure is thermally unstable. In fact, this range is often the lowest mechanical property for heat treatable aluminum alloys [17].

Fig.4. A diagram of the weld zone with several areas..

Figure .5: friction stir welding AZ-31B

Figure 6: Various tool pin geometries used for making joints.

Figure 7: Dimensions of tensile specimen[18]

Table 1: Chemistry of ZE-41 Mg alloy used

Table 2: Mechanical behavior of parent material ZE-51.

Table 3: Process parameters used

Figure 8: Specimens for tensile testing.

Table 4: ANOVA for developed model [21].

Table 5: Results from ANOVA test for UTS [22]

Figure 9(a) Material clamping_specimen

Figure 9(b) Material holding under 2 jaws

3. THERMAL MODELING OF FSW

---

3.1. Governing Equations

All thermal models, regardless of their complexity, are based on solving the classical heat equation (equation 1). (2) is known in mathematics as a scalar partial differential equation (PDE) (Betounes, 1998; Carslaw and Yeter, 1959). With proper initial and boundary conditions, this time-dependent problem can be solved with both Lagrangian and Eulerian reference systems. In the first scenario the heat field is defined by a fixed coordinate system, whereas in the second scenario the coordinate system moves with the heat source, as shown schemat-ically in Figure 2. In the Lagrangian frame of reference, the temperature change of a region over time is ex-pressed as [23]:

where the constant component K/ρcp , is the material's thermal diusivity, which is a measure of how quickly heat is transferred in a solid.

ρ denotes the material density=kg/ m3

cp the specific heat capacity =J/kg K

T the temperature field (K)

k the thermal conductivity=W/mK

qvol the volumetric heat source term =w/ m3

A convective term is introduced to Eq. when describing heat flow in an Eulerian reference frame (2).

where u is the velocity eld vector, which can include both the tool welding velocity and the material ow eld sur-rounding the tool probe, i.e. in the shear layer.

If transitory effects are insignificant, the time- dependent component, i.e. T t, disappears, and Eq. (2.1) in the Lagrangian reference frame is simplified to

Fig.10.A diagram of the Lagrangian and Eulerian reference frames

The former strategy is relatively common, while the latter can benefit from the steady-state conditions that can occur with moving heat sources [24]. Table 2.1 provides a tabular overview of these formulations. In addition to specific numerical applications, it also provides important analytical formulas. This is discussed in detail in the next section [21]. The combination of frame of reference and interdependence of modeling layers provides a wide range of choices, and the 'right' decision depends on the purpose of the particular model. For example, residual stress models require temperature fields from transient Lagrangian thermal models [25].

Table-6.governing (heat conduction) equations in thermal models in many reference frames and temporal domains.

3.2. Application of the Thin-Plate Solution in FSW

Equation 1 shows a schematic diagram of a linear heat source moving with constant velocity along a line in an infinite plate. The same partial differential equation of heat conduction, including the convec-tive coefficient, governs the temperature field, as shown in Equation 3. Under the quasi-steady state as-sumption given by Eq., the closed solution is given by Eq.-6 [28].

where (Qtot/d) defines the heat input per unit thickness, ξ denotes the moving coordinate axis

a is the heat diffusivity, K0 denotes the Modified Bessel Function of the second kind and zeroth order and T0 defines the ambient temperature [33].

Fig.11(a) Schematic view of a line-type heat source moving along the joint line of the work piece[26]

shows the resulting temperature profile along the weld axis at various distances, i.e., y = 0, 5, 10, 15 and 20 cm from the bond line, when the heat source passes x = 0 m. increase. Note the infinite temperature mathemat-ical singularity at the heat source location. This is not possible in his real FSW application where the solidus temperature is the limit of the peak temperature [27].

Fig.11(b) Temperature profile along x-axis at different y- points

Fig.12. (a) Temperature weld on a 3 mm-thick plate

3.3. Prescribed heat source models in FSW.

Heat generation due to a combination of friction and plastic dissipation is often simulated in most purely thermal FSW models using a surface Ux boundary condition at tool/matrix contact [33]. The most important unknowns in these surface Ux equations are either the friction coefficient under the slip assumption or the material yield stress under the stick assumption [30].

Along this line, Schmidt and Hattel (2004) have proposed the following generally adopted equation for the total heat generation [W],

Fig-13. Schematic view of analytical tool geometry (including conical shoulder and cylindrical probe and heat generation contributions from shoulder Q1, probe side Q2 and probe tip Q3).

However, when implementing this into a numerical model using a position dependent surface flux [W/m2 ], it is typically used on the following form[31].

---