PSI - Issue 25

Romanin Luca et al. / Procedia Structural Integrity 25 (2020) 149–158 Author name / Structural Integrity Procedia 00 (2019) 000–000

152

4

Fig. 3. Equivalent loads applied according to the Virtual Weld Bead Method for a T joint

Before starting with the inverse analysis, global deformations of experimental points are processed to extract angular deformation, longitudinal and transverse shrinkage. The load vector is calibrated one value at a time either with analytical or stochastic approach. Taking advantage of the superposition principle, every load component is first calculated individually and then assembled in the final vector. The reference deformed shape is translated into only six scalars containing the transversal shrinkage, in millimetres, and angular deformation, in degrees, for each side of the joint. These scalars represent the target to be reached by the virtual Weld Bead method. Inverse analysis is then performed with linear elastic material following the small deformation theory. The same settings are to be used on the global model. The process is very fast thanks to the superposition principle (Fig. 4).

Fig. 4. Scheme of the superposition principle employed on each deformation mode

The major advantages of the presented simplified method could be summarised as follows:  Fast geometry pre-processing phase (pre-defined length of the virtual bead)  Mesh insensitive  Easy load application