PSI - Issue 61

Shahriar Afkhami et al. / Procedia Structural Integrity 61 (2024) 53–61 Shahriar Afkhami et al./ Structural Integrity Procedia 00 (2019) 000 – 000

55 3

was applied on their edge as joint preparation. To calculate the pre-strains caused by cold-forming, 2D-simplified non linear FE models were used. These models were developed using 2D triangular elements for meshing with 0.5 mm size and FFEPlus solver available from Dassault Systèmes (SolidWorks version 2022). As noted in the introduction, a modified joint design was used in this study compared to the authors’ previous work published in [4], and consequently, the thermal gradients in the joint area were expected to change. For comparing the temperature fields in the two designs, full-scale 3D thermal models for FE analyses were developed using the Simufact Welding simulator (version 2022.0.1). Solid elements with a size of 1 mm were used in the thermal models. Further, mesh convergence studies were carried out to ensure the reliability of FE models in both thermal and mechanical simulations. The FE models used for calculating the pre-strains (done with SolidWorks) and thermal gradients (done with Simufact) were independent of each other and not coupled. Regarding the FE model done with SolidWorks, the stress-strain curve and data of the material from [4] were used for the model. For the model done with Simufact, the thermal properties of UHSSs from Simufact ’ s material library were used to simulate the welding gradients; these properties are presented in Table 1.

Table 1. Material properties used in the FE models carried out with Simufact. Solidus limit Melting temperature

Latent heat for melting Thermal conductivity* Specific heat capacity*

Density*

1463 ℃

1510 ℃

256400 J

≈ 36 W/m.K

≈ 0.44 J/gr.K

7835 kg/m 3

* Parameter’s value at room temperature

The bent plates were welded to the straight ones using the welding parameters in Table 2. After welding, Vickers hardness was measured along the weld area using a 5 Kg load and 10 s dwell time. Next, quasi-static tensile tests were carried out using standard specimens machined out of the welded joints. The tensile specimens were 6 mm thick and were prepared per dimensions proposed by ASTM E8M for rectangular samples [10]. The tensile test was carried out three times for each joint setup to ensure the reliability of the results. The digital image correlation technique (DIC) was used during the tensile tests to identify the local strains and failure locations during the tensile tests using an ARAMIS DIC system. After the tests, fractured samples were subjected to fractography using optical and scanning electron microscopy (SEM) techniques.

Table 2. Welding parameters (average values). Pass No. Current Voltage Travel speed Travel angle

Tilt angle Filler wire

Wire feed

Shielding gas

1

230 (A) 25 (V)

10 (mm/s)

4° forehand

+20°

Union X96 11 (m/min) Ar+8% CO 2

2 & 3

220 (A) 25 (V)

10 (mm/s)

10° forehand -5°

Union X96 11 (m/min) Ar+8% CO 2

3. Results and discussion According to the mechanical FE results presented in Table 3, the plastic strain in the cold-formed area of the bent base metals changes in a range of approx. 9%-14% following their R . Further, the thermal FE results, e.g., the visual view shown in Fig. 1, indicated that the weld cross-section in the modified design used in this study, compared to the one used in [4], experienced relatively higher peak temperatures and slightly more rapid cooling rates along the fusion line and in the heat-affected areas. The difference in the thermal gradients can be attributed to the various heat sink effects caused by different joint designs due to their different geometrical features (see Fig. 1). Such variations can result in different microstructural features and HAZ softening patterns that require further investigation. Accordingly, these issues are discussed in more detail and accompanied by experimental results from the mechanical tests in the following.