PSI - Issue 45

Yuanpeng Zheng et al. / Procedia Structural Integrity 45 (2023) 96–103 Author name / Structural Integrity Procedia 00 (2019) 000–000

97

2

1. Introduction Metallic infrastructures, including steel bridges, traffic signs and pipelines, are prone to fatigue cracking. Relative to conventional stop-hole and welding, CFRP reinforcement is deemed a more desirable solution, owing to its high repair efficiency, ease of installation and no additional damage (Huang et al. 2019, 2022). Various types of steel specimens strengthened by CFRP have been tested to study the fatigue behavior (Chen et al., 2019a, 2019b; Chen and Huang, 2019). Compact-tension (CT) specimens are typically used to characterize the fatigue behavior of steel and welds, though fatigue performance investigations of CT specimens repaired by CFRP, especially numerical studies (Lesiuk et al., 2018, 2020), are relatively limited. This paper presents the numerical simulations of fatigue crack growth behavior of CT specimens repaired by carbon fiber sheets and a reference unrepaired specimen. Fatigue tests on the CFRP-repaired CT specimens are first summarized, against which parallel finite element and extended finite element models (XFEM) are then developed and validated by evaluating the stress intensity factor (SIF) and fatigue life of the specimens. The experimental and numerical results are carefully compared and discussed. Special attention has also been paid to the debonding behavior between the CFRP and steel plates. 2. Previous experimental program on CT specimens repaired by CFRP An experimental program was completed on one bare and five compact tension specimens with different CFRP repair scenarios. The width W , thickness B and initial artificial crack size a n of the standard CT specimen are 70 mm, 8 mm and 14 mm, respectively. Constant amplitude fatigue tests with a frequency of 10 Hz were conducted on the specimens, employing the beach-marking loading technique in which the base-line maximum fatigue load P max was 12.5 kN and the R -ratio was 0.1. Table 1 shows the repair parameters of each CT specimen. More details of the experimental program including the crack propagation curves and the fatigue crack growth rates (FCGR, d a /d N ) calculated by the secant method have been published in Zheng and Chen (2022) and Zheng et al. (2022).

Table 1. Details of specimens and repair parameters Specimen No. Patch configuration

CFRP material type

CFRP layer No. on each side

U-1

Unrepaired (benchmark) Single-side strengthening

N/A

N/A

R-S-2

High-modulus carbon fiber sheet (FTS-C8-30)

2 1 1 2 4

R-D-1P

CFRP plate (HM-1.4 T)

R-D-1 R-D-2 R-D-4

Double-side strengthening

High-modulus carbon fiber sheet (FTS-C8-30)

3. Numerical simulation details 3.1. Verification of the XFEM method

Employing the linear elastic fracture mechanics (LEFM) theory which is widely used in fatigue analysis in steel structures, the SIF range ( Δ K ) is the pivotal parameter for repair efficiency assessment and fatigue life prediction owing to the classic Paris law of d a /d N = C ( Δ K ) m . Traditionally, SIF is calculated employing the finite element method (FEM) which requires refined and geometrically complex mesh at the crack front to perform the contour integral calculation. Once the crack size changes, to manage the salient discontinuity of the crack, remesh is indispensable in the conventional FEM method. XFEM extends the shape function of the traditional elements by means of the Heaviside enrichment term and a crack tip enrichment term to address the discontinuity, freeing the numerical simulation from repetitious remesh. After initial meshing, SIFs of various crack lengths are computed simply by updating the dimension definition of the crack, rather than the global geometric transformation of the cracked steel member.