PSI - Issue 64

3

Author name / Structural Integrity Procedia 00 (2019) 000 – 000

M. Pedram et al. / Procedia Structural Integrity 64 (2024) 621–628

623

transferred to the thermography environment. The variations in the surface temperature of the slabs were measured using a FLIR B335 camera while the slabs were heating up or cooling down in the test room environment. Data collection for each slab spanned 180 minutes for tests starting at 0°C, 40°C, and 60°C, while tests commencing at - 20°C were extended up to 240 minutes. 2.2. Finite element analysis 2.2.1. Material properties The finite element model comprises two materials, concrete as the main material and the material representing void or styrofoam. The thermal properties of both materials are given in the Table 1. The thermal conductivity (1.504 W/mK), specific heat capacity (984.67 J/kg.K) and density (2.349e3 kg/m 3 ) of concrete were adjusted based on BS EN ISO 22007-2 transient plane heat source (hot disc) measurements (Pedram et al., 2022d, Pedram et al., 2022c, Pedram et al., 2024). The values given for polystyrene are adopted from the literature (Hiasa et al., 2018, Hiasa, 2016). T he radiation heat transfer coefficient was adjusted by multiplying the emissivity of concrete (ε=0.95) by the Stefan- Boltzmann constant (σ=5.67×10 -8 W/m 2 .K 4 )(Reddy and Gartling, 2010, Bergman et al., 2011). Hence, the radiation heat transfer coefficient was set to be εσ = 53.865×10 -9 W/m 2 .K 4 . Both concrete and polystyrene were considered as isotropic material attributes without phase change.

Table 1. Thermal properties used in the FEA Properties

Concrete

Polystyrene/air

Thermal conductivity (W/mK) Volumetric heat capacity (MJ/m³K)

1.504 2.313 2349

0.024

2.825e-4

Density (kg/ m³)

25

Specific heat capacity (J/kgK)

984.67

1130

Analysis type and boundary conditions The finite element analysis in this paper simulates the experiments of concrete slabs with subsurface defects heating up or cooling down from a constant initial temperature through a transient process. During data collection, the slabs were in insulation boxes with five faces covered with polystyrene and plywood. In addition, the faces of slabs with the void/ defect hidden underneath were exchanging heat with ambient through convection and radiation. Therefore, the five faces of slabs covered with the insulation boxes were modelled with a constant heat flux (f). Because, in practice, it was impossible to create a zero heat flux condition on the insulated sides due to technical limitations. The surfaces of models, with voids underneath, were modelled as convection and radiation surfaces. Fig. 1 demonstrates the boundary conditions for this problem. The nine points on the surface are the points adopted for temperature analysis from the sequence of IR images collected during experiments. By considering the radiation coefficient, the nonlinear solver had to be selected in the LUSAS software package. The analysis was performed in two stages (load cases). The purpose of the first load case was to adjust the initial temperature and the purpose of the second load case was to achieve the transient solution. The second load case was solved in 5-minute time steps to match the 5-minute rate of IRT experiments.