PSI - Issue 59

S.E. Donets et al. / Procedia Structural Integrity 59 (2024) 367–371 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

369

3

Fig. 2. Experimental stand to heat the water pipe: (right) optical image of the set up and (left) thermal image at 773 s after the start of experiment (cooling). The respective imaging area (right) is highlighted with the rectangle on (left).

Fractal character of the thermograms was studied as per methodology described in Donets (2023). Fractal dimension analyses were conducted using the in- house software package based on the classic Clarke‟s triangular prism surface area method (TPSAM) on the squared windows 8x8 pixels with a sliding step of 1 px, basics of that algorithm were described in Clarke (1986). An image is considered as a matrix of certain x×y pixel size, while temperature of a thermal image is considered as an intensity, Rathore (2014). Having a surface in such 3 dimensions, it is possible to calculate a fractal dimension map. Regarding a relative error of calculations of fractal dimensions, it was estimated from actual tests on images to be around 0.01. The visualizations and data post-processing were performed using Python 3 in the Jupyter notebooks. Microhardness testing was done using a PMT-3 tester. 3. Results Major damage & pollutant sources were found to be sediment buildup and encrustation. Rust, mineral deposits incl. limescale were noticeable in the internal surface. At a microscale, the surface had numeral cavities and signs of pitting corrosion on the internal surface. The cold-water supply appeared to be an aggressive environment leading to severe operational degradation of the pipe. At the start of experiment, when cold water started to flow through the experimental stage, no temperature heterogeneities were visible. Once the heating was applied, surface temperature monitored on the external surface started to vary. An example of a thermogram with the heterogeneous distribution is presented on Fig. 2. Basically, presence of polluting objects and damage on the internal wall of the pipe impact the water flow which in its turn lead to differences in thermal flows. As the wall of the pipe is relatively thin, the temperature differences become visible within few seconds at high temperature gradients. Thermal heterogeneities were noticeable at gradients > 0.04 K/s, however more preferable conditions were at >0.2 K/s. It is not possible to clearly distinguish minor damage (couple millimeters in diameter) or pollution unless a defect is very close to the surface as propagation of thermal flow gives a smooth view. Pitting corrosion is not visible at all as the wall thickness is too large while pits are not deep enough to penetrate the wall of the pipe. In general, active thermography in the current setup helps to localize some large heterogeneities along the length of the pipe, which significantly distort the thermal flow like clogging of sediments. As pointed in Yu (2023), it may be possible to reveal defects with residual thicknesses of up to 3 mm and diameters of up to 5 mm for steel – sand or steel – air defect interfaces at advanced laboratory conditions. On Fig. 3, there are shown linear profiles on the pipe‟s temperature representative for heating, heating plateau, and coo ling phases. Strong variations in temperature are pointing to locations of structural abnormalities or objects hindering the normal flow of the water. Once plotted with corresponding thermographic image or optical image over at different phases of heating-cooling cycle, abnormalities are localized. Regarding practical application of this method in real life, it is important to achieve relatively high temperature gradients to start seeing the structural heterogeneities. It can be done by abrupt introduction of hot or cold water in the system, or even when the water pipes were empty and rest at ambient temperature while introducing cold water into them. Also important is to have a relatively clean surface of the pipes, with no dust or moisture. When the surface is not clean, it is hard to distinguish temperature