PSI - Issue 47

Teresa Morgado et al. / Procedia Structural Integrity 47 (2023) 882–887 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

883

2

Keywords: Conventional Finite Element Method; XFEM Method; Railway component; Fatigue

1. Introduction There are two principal methodologies to obtain the SIF (Stress Intensity Factor) values numerically: one is based on analysing the state of stresses and strains near the crack tip, and the other is determining the variation of the energy levels generated in the crack advancement. The first methodology is based on determining the Integral J around an arbitrary contour, which can be easily related to the strain energy release rate, G and consequently SIF, assuming the Linear Elastic Fracture Mechanics (LEFM) conditions (Shih et al., 1986), (Kobayashi et al., 1973 ), (Anderson, 2005), (Zhu et al., 2012), (Xue et al., 2013). This study is carried out by numerical simulation through analysing solutions of the J-integral and the Stress Intensity Factor (SIF) in a specimen subjected to a three-point bending test (Single Edge Notch Bending Test – SENB test). The material studied is ASTM A148 90-60 used in railway components (Infante et al., 2003), (Morgado, 2016). The research approach used in developing this work consisted of dividing the study into two parts; 2D and 3D Numerical Simulation Studies. The 2D numerical simulation was subdivided into six studies, as shown in Table 1. And the 3D numerical simulation was subdivided into two studies, presented in table 2. In numerical simulation has used the software Abaqus CAE and two techniques were applied to evaluate the SIF and Integral J solutions: the Conventional Finite element Method and the Extended Finite Element Method (XFEM) (Abaqus,2014).

Table 1. Study WĂƌƚ ϭ͗ Ϯ EƵŵĞƌŝĐĂů ^ŝŵƵůĂƚŝŽŶ ^ƚƵĚLJ

Study

Method

Objectives

Define crack length. Define element type.

➢ ➢

1)

Conventional

2)

Conventional

Influence of mesh refinement.

➢

Validation of the model without contacts. Define the methodology to use in the fatigue study.

➢ ➢

3)

Conventional

4)

XFEM

➢ Influence of mesh refinement in the crack propagation. ➢ D efine the values of ∆SIF a nd comparate them with experimental ∆K obtained by Morgado (Morgado, 2015). ➢ Validation of the method in the fatigue crack propagation. ➢ Comparison of the crack length obtained by simulation with the crack length obtained by analytical method.

Conventional with Fatigue

5)

XFEM with Fatigue

6)

Table 2. Study WĂƌƚ Ϯ͗ ϯ EƵŵĞƌŝĐĂů ^ŝŵƵůĂƚŝŽŶ ^ƚƵĚLJ Study Method

Objectives

Determine the influence of refinement in the results obtained.

➢ ➢

1)

Conventional

Compare the results with 2D.

2)

XFEM

Comparison of 2D and 3D results.

➢

2. Numerical simulation Fig. 1 presents the specimen dimension used in numerical simulations as in the experimental test (Morgado, 2015), where: L = 100mm, S=80mm, W=20mm and B=10mm.